Army STTR ARM26TX06-NV003: Agentic-AI, Schema-Driven Decision Management for Auditable Studies and Acquisition Decisions
Quick Answer
ARM26TX06-NV003 is a U.S. Army STTR Phase I topic released under the Department of the Army 2026 STTR Program Commercial Solutions Opening (CSO), Release 6. The Army is asking small businesses, working in formal partnership with a research institution, to prove out a "decision-as-a-program" capability: an agentic AI system built on a formal decision schema that turns messy engineering and acquisition questions into structured, auditable, signer-ready Decision Packages. Phase I awards are up to $300,000 for a period of performance of 1 to 6 months. The topic opens September 23, 2026 and closes October 21, 2026, with proposals certified in the Defense SBIR/STTR Innovation Portal (DSIP) no later than 12:00 p.m. ET on the close date.
The single most important thing to understand about this topic is that it is the STTR version of an Army topic that also exists as an SBIR topic under number ARM26BX06-NV012. The technical scope, the funding ceiling, the period of performance, and the evaluation rubric are the same. What differs is the program: STTR requires you to partner with a research institution and to divide the work with that partner according to statutory minimums. If you do not have a university or federal laboratory partner lined up, the SBIR twin is the door you want. If you do have one, and especially if the principal investigator sits on the academic side, this is the better fit.
Topic At a Glance
Topic number: ARM26TX06-NV003
Title: Agentic-AI, Schema-Driven Decision Management for Auditable Studies and Acquisition Decisions
Program: STTR, Small Business Technology Transfer
Solicitation: Department of the Army 2026 STTR Program, Commercial Solutions Opening (CSO), Component-Specific Proposal Instructions, Release 6
Agency: U.S. Army, under the Department of War (DoW) STTR program
Phase offered: Phase I (Feasibility) only
Award ceiling: Up to $300,000
Period of performance: 1 to 6 months
Contract type: FAR-based, firm-fixed-price, with payments tied to measurable milestones or deliverables
Research institution partner: Required, as a condition of the STTR program
OUSD (R&E) Critical Technology Area: Applied Artificial Intelligence (AAI)
Component Technology Priority Areas: Trusted AI and Autonomy; Information Systems
Projected CMMC level requirement: Level 1
Submission portal: DSIP at dodsbirsttr.mil
Topic open date: September 23, 2026
Proposal deadline: October 21, 2026 at 12:00 p.m. ET
DSIP Topic Q&A closes: October 7, 2026, two weeks prior to the close date
Selection notification: Within 90 days of the topic close date, meaning on or about January 19, 2027
SBIR companion topic: ARM26BX06-NV012, same title and scope, no research institution partner required
Keywords: Agentic AI, Digital Engineering, schema-driven Decision Management, acquisition decisions, AI LLM
A note on the topic number. The Release 6 topic index lists this topic as ARM26TX06-NV003, while the header on the topic description page carries a typographical variant. Use the number shown on the live DSIP topic posting, and confirm it before you submit, since the topic number goes in your technical volume header on every page.
What Makes This an STTR and Not an SBIR
This is the part worth getting right before you invest any writing time.
STTR is a distinct program from SBIR, authorized under the same statute but with different structural requirements. The defining requirement is the research institution partnership. An STTR award is a collaboration between a small business concern and a research institution, typically a university, a federally funded research and development center, or a qualifying nonprofit research organization. The partnership is not optional and it is not a subcontract of convenience. It is the point of the program.
Two consequences follow, and both come from the STTR statute and the SBA SBIR/STTR Policy Directive rather than from the Army component instructions themselves. First, the work must be divided between the small business and the research institution according to statutory minimums, with the small business performing at least 40 percent of the work and the research institution performing at least 30 percent, measured as a percentage of the total effort. Second, the principal investigator may be primarily employed by either the small business or the research institution. That second point is spelled out in the Army instructions, which state that the primary employment of the Principal Investigator or Project Manager must be with the small business concern, or with the research institution in the case of STTR, at the time of award and during the conduct of the project.
Because the exact terms, definitions, and the required written allocation-of-rights agreement live in the DoW 2026 STTR Program CSO and the SBA Policy Directive rather than in the Army component-specific instructions, read those documents alongside this topic. The Army instructions are explicit that they supplement, and in some cases supersede, the DoW CSO, but they do not restate the STTR partnership mechanics.
One eligibility difference is worth flagging carefully. The Army SBIR component instructions for Release 6 include a section addressing small businesses that are majority-owned in part by multiple venture capital operating companies, hedge funds, or private equity funds, confirming those firms are eligible to submit and receive SBIR awards provided they complete a joint venture eligibility certification. The Army STTR component instructions for the same release do not contain that section. The majority venture-capital ownership authority is an SBIR-specific provision of the statute. If your company is majority-owned by multiple VC, hedge fund, or private equity firms, verify your STTR eligibility against 13 CFR 121.702 and the SBA SBIR/STTR Policy Directive before you commit to this topic, because your answer under STTR may differ from your answer under SBIR. Note that the Verification of Eligibility of Small Business Joint Ventures document still appears on the STTR conditional Volume 5 list.
If you are choosing between the two topics, the practical decision tree is short. No research institution partner means go SBIR. A strong university or laboratory partner whose people and instruments you actually need, or a principal investigator whose primary employment is academic, means go STTR. A partner you would be adding only to qualify for STTR means go SBIR, because a thin partnership shows up in the Technical Team and Technical Approach scoring.
What the Army Is Actually Looking For
The technical content of this topic is identical to its SBIR twin. Everything below comes from the topic description in the Release 6 STTR instructions.
The core objective
The Army wants proof of a decision-as-a-program capability. In plain terms, that means taking a complex engineering or acquisition question, one that today gets answered through slide decks, spreadsheets, email threads, and tribal knowledge, and turning it into a structured, machine-readable, auditable decision artifact that a decision authority can actually sign. The stated goal is to do this faster and with higher confidence than the status quo.
The emphasis is on proof by demonstration, not a full product build. Phase I is not the place to promise a shipping platform. It is the place to show that the schema plus the agentic layer works on real decision problems, both point-in-time trade studies and long-horizon acquisition decision programs, with traceability, robustness, and human control intact.
The three technical objectives
Objective one: a unified representation. The Army wants a single decision model that supports two very different shapes of work. The first is the single-episode trade study, a bounded comparison of options against criteria. The second is the multi-episode "decision program," a long-running decision with refresh cycles as evidence and assumptions change over time. The topic specifically calls for context engineering for the ground vehicle domain, implemented through knowledge graphs and semantic layers that capture domain-specific relationships, constraints, and requirements. The system is expected to integrate data from Digital Engineering (DE) tools such as SysML 2.0 and UML, pulling from digital threads and digital twins to provide a unified view of system models across the product lifecycle.
Objective two: agentic capabilities with human control. The AI layer must produce structured artifacts and reproducible outputs. Reproducibility is called out explicitly, which is a strong signal that nondeterministic, unauditable model output alone will not satisfy this topic. The agentic system is expected to perform structured elicitation into the formal model, generate decision workflow plans, and execute evaluation runs with clear "what flips the decision" logic. It is also expected to automate validation of technical documents, models, and outputs against standards and requirements, while maintaining predictable behavior consistent with established human-AI interaction guidance.
Objective three: explicit bias checks, risks, and assumptions. The schema must be extended so that bias checks, risks, and assumptions become first-class, testable objects rather than prose buried in a report. The purpose is decision readiness: the ability to assess whether a decision is actually ready for sign-off before it is signed, and to responsibly refresh it when conditions change. Semantic understanding is expected to keep consistency across digital engineering artifacts and enable traceable decision rationale.
Phase I desired outcomes
The topic lists three Phase I outcomes:
A decision schema that makes objectives, options, constraints, assumptions, risks, and bias checks first-class objects.
Agentic AI that performs structured elicitation into the formal model, generates decision workflow plans, and executes reproducible evaluation runs with clear "what flips the decision" logic.
Two end-to-end demonstrations showing the approach scales from point-in-time trade studies to long-horizon program decisions with refresh cycles as evidence and assumptions change.
Two demonstrations, not one. That is a concrete, checkable requirement and a common place where proposals lose points by being vague.
Where this goes in Phase II
The Army is explicit that the DoW already has multiple contractors building AI and LLM tools for core warfighter purposes, but very little effort has gone toward an AIgentic decision intelligence capability that supports accelerated commercial acquisition or longer lead traditional development. That gap is the reason this topic exists.
In Phase II, the expectation is that the successful contractor develops a combined AIgentic and AI/LLM decision intelligence workflow capability, exercised on two real cases provided from PAE Maneuver Ground and PAE Fires. The resulting software platform is to be modeled and developed as a commercial open-source platform usable by both the DoW and commercial enterprises for Product Design and Development (PDD) and acquisition.
The topic names the current COTS baseline directly: DAOSoft, which provides an open capability with an advanced structured decision and trade study capability plus a partially developed AI/LLM decision intelligence workflow. The Army states it is looking for any and all ideas that can extend or replace current acquisition capabilities and is agnostic to all current solutions. It is also agnostic as to which developed AI commercial package fulfills the addition of AIgentics to current AI/LLM for continuous acquisition.
The Phase II prototype is expected to be immediately useful, with follow-on Phase III contracts anticipated for individual customer installation, described by the Army as similar to current best practice deployment of large ERP systems from SAP and Oracle.
For an STTR team, the open-source and platform expectations in Phase II deserve an early conversation with your research institution's technology transfer office. Who owns what, what gets published, and what can be released under an open-source license are questions that get much harder to answer after award than before.
Phase III dual use
The named commercial analog is the vehicle design process in the automotive industry. The Army notes the workflow can be relatively similar to the vehicle design process for the Army, with some different considerations. For a startup, this is a useful signal: the commercialization argument writes itself around complex-system product development and capital program decision-making outside defense.
Reference cited in the topic
The topic cites one reference, an SAE Mobilus paper: "Concept execution AI agentic decision intelligence framework product planning concept development," 2025-01-0455, at saemobilus.sae.org.
Funding Allowance and What You Can Add On Top
Base award
Phase I under this topic is up to $300,000 for 1 to 6 months. Award funding guideline and period of performance limitations are established per topic, and the solicitation is blunt about this: proposals exceeding these limitations will be deemed unresponsive. Do not propose $325,000 and do not propose nine months.
Remember that this single ceiling covers both partners. The $300,000 is the total award, split between your small business and your research institution according to the STTR work allocation. That constrains scope more than it does under SBIR, because you are funding two organizations with two different indirect rate structures out of the same pool. Build the budget early and check it against the statutory work split before you finalize the technical approach.
The number of awards depends on funds availability. The Army reserves the right to award none, one, or more than one contract under any topic. Awards are FAR-based, firm-fixed-price, with payments tied to measurable milestones or deliverables agreed to by the Government. Milestone schedules are used to monitor technical progress, mitigate technical and cost risk, and address awardee cash flow needs. The Contracting Officer retains the right to negotiate contract type and price.
Technical and Business Assistance (TABA)
TABA is additional money on top of the award ceiling. For Phase I, the Army may provide up to $6,500 per project. For Direct to Phase II projects, up to $50,000. TABA is not guaranteed and depends on the Government's review of the requested services and available TABA budget.
There are two options and the paperwork differs:
Option 1, Army-preferred vendor. Up to $6,500 per Phase I project. If you use the Army's preferred provider, the TABA cost does not need to be included in your DSIP Cost Volume (Volume 3). If approved, funding goes directly to the Army's preferred provider. Participation is voluntary and application instructions come after contract award. Commonly requested services include customized market research and strategy, intellectual property support, CMMC compliance, DCAA compliance, and communications and marketing. Intellectual property support is worth a look for an STTR team specifically, given the allocation-of-rights work that comes with a research institution partnership.
Option 2, firm-selected vendor. You may contract TABA services through one or more TABA providers, or use the funding to hire new or augment staff, or to have staff conduct or participate in training activities consistent with 15 U.S.C. 638(q)(1). If you choose your own provider, the proposed TABA amount must be included in the DSIP Cost Volume (Volume 3) and details of the TABA request must be included in Supporting Documents (Volume 5).
Critical TABA compliance points:
TABA funding will be denied if the costs are omitted from the initial proposal submission, or if proposed TABA costs exceed $6,500 per Phase I project.
TABA costs cannot include any profit or fee for the proposer.
TABA costs cannot be factored into the calculation of general and administrative expenses (G&A).
The amount is based on the original contract period of performance and does not apply to period of performance extensions or enhancements.
The TABA provider cannot be your firm or any affiliated entity, and TABA costs must be directly related to the provider's services, excluding indirect costs, profit, or fees passed through by your company.
The rationale statement in Volume 5 is limited to two pages and must include the provider name, point of contact with email and phone, an explanation of the provider's suitability and qualifications, an outline of specific tasks, and total provider cost with number of hours and labor rates. A blended or average rate is acceptable.
Phase II Enhancement
The Department of the Army may provide a Phase II awardee with up to $500,000 in matching SBIR funding on a dollar-for-dollar basis if the performer obtains a commitment of non-SBIR/STTR funding from a DoW component, another federal agency, or a commercial investor. The enhancement effort must develop, deliver, and integrate a technology or product into a program within a DoW component, federal agency, or the commercial sector. Enhancement requests are subject to approval by the Army SBIR/STTR Program's Source Selection Authority, successful negotiations, and funding availability. Enhancement funding is typically applied to an active Phase II award through a contract modification, with additional scope of work and period of performance commensurate with the total funding received.
For a startup, this matters at the strategy level: a credible non-dilutive path exists from a $300,000 Phase I to a Phase II with up to $500,000 in matched enhancement, then Phase III installation contracts.
Timeline and Key Dates
The dates that matter
Topic opens for submission: September 23, 2026
DSIP Topic Q&A closes: October 7, 2026
Deadline for administrative questions to the Army: October 6, 2026, which is 15 days prior to closing
Proposal deadline: October 21, 2026 at 12:00 p.m. ET
Selection or non-selection notification: on or about January 19, 2027, within 90 days of close
Start of Work Meeting, if awarded: within 30 calendar days after contract award
Note that the proposal deadline is noon Eastern, not end of day. That is 11:00 a.m. Central, 10:00 a.m. Mountain, and 9:00 a.m. Pacific on October 21. Teams on the West Coast routinely lose proposals to this detail.
The sequence in detail
Pre-release period, before September 23, 2026. The topic sits in pre-release on DSIP. During pre-release, topic author names, phone numbers, and email addresses are published with the topic on the DSIP Topics and Topic Q&A page. This is the only window for direct contact with the topic author, and once the topic opens on September 23 that door closes.
Topic open, September 23, 2026. The topic opens for proposal submission in DSIP.
DSIP Topic Q&A window, closing October 7, 2026. Topic Q&A opens on the pre-release date and closes two weeks prior to the topic close date. All questions and answers are posted publicly on a non-attribution basis. Answers are generally posted within seven business days and are also emailed directly to the person who asked. Because answers can take seven business days, a question submitted on October 7 may not be answered until roughly October 16, leaving you only a few days to act on it. Ask early.
Administrative question deadline, October 6, 2026. General questions about Army STTR Program administration and the Army component-specific proposal instructions should be submitted as soon as possible, and no later than 15 days prior to solicitation closing, to usarmy.SBIRSTTR@army.mil.
Proposal deadline, October 21, 2026 at 12:00 p.m. ET. Complete proposals must be certified and submitted in DSIP by this time. Proposals cannot be submitted after the deadline and will not be accepted or evaluated.
Selection or non-selection notification, on or about January 19, 2027. Within 90 days of the topic close date, by email to the Corporate Official listed on the proposal cover sheet, sent from the Army SBIR/STTR Program Office mailbox.
Start of Work Meeting, within 30 calendar days after contract award. Held at your facility unless another location is designated in the contract.
Periodic reviews. Monthly, bi-monthly, or quarterly review meetings as specified in the contract.
Final Report of Inventions and Subcontracts. DD Form 882, due within six months after completion of the contracted work.
Planning note: the Army warns that proposals should be submitted as early as possible to avoid delays caused by high traffic near closing. The Department of the Army is not responsible for missed submissions due to system latency or inaccessibility. A proposal sitting in "In Progress" or "Ready to Certify" status is not submitted, even if every volume has been uploaded before the close date.
A working backward schedule for an STTR team
The 29 day window is tighter for STTR than for SBIR, because a second institution has to move with you. University contracting offices do not work on startup timelines.
Now through September 22. Contact your research institution's sponsored programs office immediately and tell them the deadline. Ask what they need from you and how long their internal review takes, because that answer sets your real deadline, not October 21. Confirm your SAM registration is active and that your DSIP profile matches it. Get every team member a DSIP account. Complete the annual Volume 6 Fraud, Waste and Abuse training. Reach the topic author during pre-release while direct contact is still allowed.
September 23 through September 30. Settle the work split with your partner on paper, in percentages, and confirm it satisfies the STTR minimums. Get the research institution's budget and indirect rates in writing. Run Army customer discovery interviews. Draft the technical approach and lock down which two demonstrations you will propose.
October 1 through October 7. Complete a full technical volume draft. Submit any remaining questions through DSIP Topic Q&A before it closes on October 7. Circulate the draft to your research institution partner for their sign-off on anything describing their scope, their people, or their facilities.
October 8 through October 14. Build the combined Cost Volume with substantiating documentation from both organizations. Assemble every Volume 5 attachment, including the two page Non-Proprietary Work Plan, the foreign citizens list or the statement that none are involved, and Key Personnel with the principal investigator's technical resume. Complete the Volume 7 foreign affiliations web form. If your research institution requires institutional signature on anything, this is the week it has to happen.
October 15 through October 17. Compliance pass. Verify the seven page limit, 10-point minimum type, one inch margins, page numbering, and headers with company name, topic number, and DSIP proposal number. Confirm the PDF is searchable, unlocked, unencrypted, and virus checked.
October 18 through October 19. Upload everything to DSIP and execute the corporate official certification. Confirm the status is no longer "In Progress" or "Ready to Certify."
October 20 through October 21. Buffer only. Do not plan to use these days.
Eligibility and Registrations You Need Before You Can Submit
You must be a Small Business Concern (SBC) meeting all eligibility requirements in 13 CFR 121.702 and the SBA SBIR/STTR Policy Directive, and you must have a qualifying research institution partner as required by the STTR program.
System for Award Management (SAM). You must be registered and active in SAM in accordance with FAR Provision 52.204-7, both when submitting and at time of award. Proposals from firms failing this may be deemed unresponsive. You may only submit using your legal business name or a DBA name as indicated in the SAM registration for the provided Unique Entity Identifier (UEI). A DBA used on an offer must be registered and linked to your current, active SAM registration. Firms that are debarred, suspended, proposed for debarment, or voluntarily excluded are excluded from award, and the Army will not consent to subcontracts with such contractors.
DSIP registration. DSIP is the official portal. Proposals submitted by any other means will be disregarded. Every individual who will prepare or submit must have an individual DSIP user account. Register as early as possible. Your DSIP account profile must correlate with your SAM registration, including the 5-digit CAGE code, 12-digit UEI, legal business name, DBA name, and physical address. Failure to correlate this information, or to submit required certifications, may significantly delay funding agreement award or become grounds for cancellation of the funding agreement.
Principal Investigator employment. The primary employment of the Principal Investigator or Project Manager must be with the SBC, or with the Research Institution in the case of STTR, at the time of award and during the conduct of the proposed project. This flexibility is one of the real advantages of the STTR path for teams built around an academic investigator.
Research institution partnership documents. The written agreement between the small business and the research institution, including the allocation of intellectual property rights, is governed by the DoW 2026 STTR Program CSO and the SBA SBIR/STTR Policy Directive. Confirm what has to be in place at proposal time versus at award, because the answer drives your schedule.
Proposal Structure: What You Actually Submit
Army component-specific instructions supplement and in some cases supersede the DoW 2026 STTR Program CSO. Where they conflict, the Army instructions take precedence.
Volume 1, Proposal Coversheet
Prepared in DSIP in accordance with the DoW solicitation.
Volume 2, Technical Volume
Format requirements, and the solicitation states that submissions failing to conform shall be deemed unresponsive:
Single Adobe Acrobat PDF, searchable text format, including graphics. PDFs that cannot be opened using Adobe Acrobat products may be rejected. Perform a virus check before uploading. Do not lock, password protect, or encrypt the file. Do not include or embed active graphics such as videos or moving pictures.
Page limit for Phase I: seven pages, including all key sections. The limit applies to the entire Technical Volume as uploaded to DSIP. At the Government's discretion, pages that clearly do not substantiate or contribute to the technical approach, such as a cover sheet or table of contents, may be excluded from the count. Any appendices or supplemental content that expand upon the required content do count toward the seven pages. You may allocate the pages across sections however you want.
Layout: number all pages consecutively. No type smaller than 10-point, standard 8.5 by 11 inch paper, one-inch margins including the header. The header on each page, which may sit inside the one-inch margin, should contain the proposing SBC's name, the topic number, and the DSIP assigned proposal number from the cover sheet.
General: graphics are strongly encouraged throughout. For plots and charts, include title, caption, axes labels, and scale. Avoid jargon and define all technical terms. Use the compress pictures feature to reduce file size.
Required section order for Phase I:
Introduction
Army Benefits
Technical Approach
Programmatic Potential
Commercial Potential
These sections map one-to-one onto the Phase I evaluation rubric in Appendix A. Your Volume 2 must also be consistent with the Non-Proprietary Work Plan you upload in Volume 5.
Seven pages is not much room to describe a two-organization project. The research institution's role has to be established economically, inside the Technical Approach and Technical Team narrative, and it has to read as integral rather than bolted on. Detailed resumes belong in Volume 5, not here.
What each section needs, in the Army's own framing:
Introduction. A clear, concise description of what your innovation does or will do, and where you are in your evolution. Make clear its intended impact on the Army. Evaluators should "get it" after reading this.
Army Benefits. Identified Army use cases, the solution's advantages, and the potential level and scale of impact. If you are unsure how an Army end user would benefit, provide an analogous use case so evaluators can connect the dots. Three rubric prompts: Alignment, argue your innovation aligns with this topic's priorities as defined in the solicitation. Solution's Advantages, prove customers will choose you given limited resources and many choices, and account for indirect substitutes as well as direct competitors. Solution's Impact, the Army seeks higher-risk, high-impact solutions through STTR, not engineering changes or incremental improvements.
Technical Approach. Details and supporting data on how you will solve the problem, key elements of your approach, the technical team, and risks with mitigation plans. Use data to substantiate that your risk mitigation plans are credible, attributed to reliable, credible sources. Rubric prompts: Scientific Feasibility, Enabling Technologies (do the required enabling technologies introduce added risk), Technical Team, Technical Risk and Mitigation Plans, and Data Quality.
Programmatic Potential. Where your company is today, what will be accomplished under this STTR effort with Army customer discovery, and the risks and mitigations for transitioning beyond STTR funding into a contract with an Army or integration with an Army system. Rubric prompts: Project Milestone Schedule, Army Customer Discovery and Validation (are you "getting out of the building" and do you have formal or informal customer validation), and Army Transition Pathways (what is the next type of deal you aim to make with the Army after this award).
Commercial Potential. The commercial market the DoW can build upon, your past success, and future potential. Rubric prompts: R&D to Product Revenue (has your team transitioned R&D into products with actual product revenue, which the Army defines as directly selling a solution rather than selling consulting, services, or research activities), Competitive Edge (why will you win, something your team does very well that is difficult to match, such as protected IP, unmatched relevant experience, a novel business model, or network effects), and Other People's Money (the non-DoW commercial market case).
Volume 3, Cost Volume
Cost and duration limits are detailed in the topic. Options are not anticipated at this time; if an option is identified in the topic posting, Base and Option costs must be separated and clearly identified. All proposed costs must be accompanied by documentation substantiating how the cost was derived. Failure to include supporting documentation may delay award, since the Contracting Officer will have to request it. Failure or refusal to provide documentation may result in dissolution of the contract action.
For STTR, the cost volume carries the work split. Your research institution's costs, rates, and effort have to be substantiated the same way yours are, and the resulting allocation has to satisfy the STTR minimums. Ask your partner's sponsored programs office for a formal budget with their federally negotiated rates rather than an estimate.
Volume 4, Company Commercialization Report
Required for prior SBIR/STTR awardees. Information in the CCR will be considered during proposal evaluation.
Volume 5, Supporting Documents
This is where proposals most often die on compliance. The solicitation states plainly that a completed proposal submission in DSIP does not indicate that the mandatory supporting documents have been uploaded. Failure to upload any mandatory document as a separate attachment will result in the proposal being deemed administratively non-compliant and unresponsive. Unresponsive proposals are un-awardable and may be excluded from technical evaluation entirely.
Mandatory items, uploaded as separate attachments:
Non-Proprietary Work Plan. Two page limit. A summary based on the Volume 2 technical approach, outlining technical objectives, key milestones, and deliverables. It serves as the basis for performance requirements in the final contract and must be written for a public, technically literate audience with no sensitive or proprietary information.
List of Foreign Citizens. Identify any foreign nationals or dual citizens expected to be involved as a direct employee, subcontractor, consultant, or Research Institution. Provide a single PDF with a listing of citizenships, countries of origin, type of visa or work permit, and an explanation of anticipated level of involvement. If no foreign nationals will be involved, substitute a statement to that effect. The Army specifically warns that small businesses often incorrectly assume everyone with dual citizenship or a work permit is automatically permitted to work on SBIR/STTR projects and fail to identify them. Failure to provide this may significantly delay selection or award.
This item deserves extra attention on an STTR proposal. Research institution personnel are named in the requirement, and university laboratories commonly include graduate students and postdoctoral researchers who are foreign nationals. Ask your partner for a complete roster of everyone who will touch the project, with citizenship and visa status, early enough that it does not become a last-week scramble.
Key Personnel. Identify all key personnel with information on education and experience. A technical resume for the Principal Investigator, including publications if any, must be included. Resumes for other technical staff, subcontractors, and consultants are useful and may also be included. On an STTR, this is where your research institution investigators and their publication records belong.
If applicable, also submit: SBIR/STTR Funding Agreement Certification; Verification of Eligibility of Small Business Joint Ventures; Assertion of use, release, or disclosure restriction per DFARS 252.227-7017; DD Form 2345 Military Critical Technical Data Agreement for ITAR-subject actions; SBC-selected details of TABA request; and other documentation such as letters of support. The Army may also accept cost/pricing information and other items only as specified in the topic. Anything else submitted will be disregarded, including promotional and non-project material.
The assertion of use, release, or disclosure restriction is worth real thought on an STTR, since data rights questions get more complicated when a university's prior work and background IP are in the mix. A letter of support from your research institution partner is also an appropriate use of the "other documentation" slot.
Volume 6, Fraud, Waste and Abuse Training
Material is in the Volume 6 section of the DSIP proposal submission module and must be thoroughly reviewed once per year to proceed with proposal submission.
Volume 7, Disclosures of Foreign Affiliations or Relationships to Foreign Countries
A web form. PDF uploads are no longer accepted. Disclosure does not mean automatic denial, but failure to complete the form completely and accurately may render the proposal administratively non-compliant and unresponsive.
Per 15 U.S.C. 638(vv), 15 U.S.C. 638(g)(13), and the SBA SBIR/STTR Policy Directive, disclosures must consider subcontractors, subsidiaries, joint ventures, consultants, and their employees across four categories: foreign affiliations, including any subsidiary, parent company, or joint venture based in, funded by, or with a foreign affiliation with any foreign country of concern; contractual or financial obligations with a foreign entity, and the Army notes that a subcontracting agreement is by definition a contractual and financial obligation; related entities, meaning any foreign business entity, offshore entity, or entity outside the United States related to the SBC; and covered individuals, meaning anyone who contributes in a substantive, meaningful way to the scientific development or execution of the project, regardless of title or Key Personnel designation, whether employed by the prime, a subcontractor, a consultant, or a partnering Research Institution.
That last clause names partnering research institutions explicitly. Your partner's foreign collaborations, foreign funding, and foreign-affiliated personnel are inside the scope of what the Army will assess. Have that conversation with your partner before you fill out the form, not after.
How Your Proposal Will Be Evaluated
The process
Initial screening for responsiveness, timeliness, and SBC eligibility. Responsiveness means failing any material requirement of the solicitation, including compulsory terms and conditions, makes the proposal unresponsive. Timeliness means received on or before the due date and prior to the established set time. Assessment of responsiveness and eligibility may continue during technical evaluation and even after selection.
National security and foreign risk assessment. The Army conducts a foreign risk evaluation review of every proposal, referencing the 2026 Department of War SBIR/STTR Foreign Risk Evaluation Program Common Decision Matrix dated 08 July 2026. The Army states it relies on statutory authority rather than non-regulatory guidance documents, and that the matrix is illustrative and non-exhaustive. The Army's assessment is not limited to the factors or countries listed. It evaluates the Volume 7 Foreign Disclosure and all areas identified under 15 U.S.C. 638(vv), including cybersecurity practices, patent analysis, employee analysis including subcontractor employee foreign affiliations, foreign ownership, financial ties and obligations, and employee ties to foreign entities. If risks are identified, the Army decides at its sole discretion and case by case whether they can be adequately mitigated. The Army is not required to coordinate with the proposer to develop a mitigation plan and is not obligated to disclose specific findings. Where risks cannot be effectively mitigated, the proposal may be removed from consideration regardless of technical evaluation results.
Technical evaluation. Proposals passing initial screening receive a technical evaluation against the criteria in Appendix A, reviewed by a cadre of subject matter expert evaluators using the weighted rubric. Proposals are assessed on their own individual merit against the stated criteria, not against each other. Because foreign risk assessment runs in parallel, the Army reserves the right to pause, suspend, or entirely forego technical evaluation, and pausing does not create any right to a mitigation period. Designated support contractors, bound by non-disclosure agreements, may review submissions for evaluation purposes; submitting a response constitutes assent to that.
Other assessment considerations. Under RFO FAR 25.502, the Government will eliminate any offeror that is unacceptable for reasons other than price, including use of prohibited sources. Proposers and subcontractors are strictly prohibited from using products or services listed in RFO FAR 40.202, including but not limited to TikTok/ByteDance, Kaspersky, and specific unmanned aircraft systems.
Selection. Notification within 90 days of topic close. Selected proposals are not guaranteed a contract award, and the selection notice is not an authorization to commit or expend funds. After selection, proposals go to a Contracting Officer for further evaluation and negotiation, which may include representations and certifications, certified or other than certified cost data, and other information. Award happens only when a Contracting Officer signs an award document.
Feedback. The Army will provide feedback to offerors per the SBA SBIR/STTR Policy Directive, with instructions in the selection decision notice. Offerors are entitled to no more than one feedback per proposal. This is not a FAR Part 15 debriefing, and offerors are neither entitled to nor will receive FAR Part 15 debriefs.
Phase I scoring weights
The Phase I Evaluation Criteria v4-0 rubric in Appendix A assigns these weights:
Introduction, 3 percent
Army Benefits, 15 percent, covering Alignment, Solution's Advantages, and Solution's Impact
Technical Approach, 35 percent, covering Scientific Feasibility, Enabling Technologies, Technical Team, Technical Risks and Mitigation Plans, and Data Quality
Programmatic Potential, 20 percent, covering Project Milestone Schedule, Army Customer Discovery and Validation, and Army Transition Pathway
Commercial Potential, 25 percent, covering R&D to Product Revenue, Competitive Edge, and Other People's Money
Proposal Quality, 2 percent, covering Quality of Prose and Images, Charts, Graphics
Each element is scored Unsatisfactory, Marginal, Satisfactory, or Superior. A few of the distinctions worth internalizing, because they tell you exactly what "Superior" looks like:
Alignment: Superior is "perfectly aligned with this Army topic's priorities," not merely "aligned."
Solution's Impact: Satisfactory is "significant improvement vs. the state of the art." Superior is "radical improvement."
Enabling Technologies: Satisfactory "requires proven technologies." Superior "requires Army-fielded technologies." Marginal "requires emerging, cutting edge technology," and Unsatisfactory "requires nonexistent or unavailable technology." This is a real trap for AI teams. Building on exotic, unproven infrastructure scores you down, not up.
Technical Team: Superior is "technical people are highly qualified AND have significant experience." Note the conjunction. Satisfactory accepts highly qualified or significantly experienced. This is the criterion where a genuine research institution partnership pays for itself, since the combination of academic credentials and applied delivery experience across two organizations is exactly what the top of this scale describes.
Army Customer Discovery and Validation: Unsatisfactory is "no customer interviews completed, no validation." Marginal is "a handful of customer interviews completed, no validation." Satisfactory is "extensive interviews completed, early validation beginning to inform transition strategy." Superior is "exhaustive interviews completed, validation informs credible transition strategy." You cannot score well here without having actually talked to Army stakeholders before you submit.
R&D to Product Revenue: Unsatisfactory is "no evidence of creating product revenue from R&D efforts." Superior is "evidence of R&D yielding product revenue sufficient to fuel this company's growth." Note again the Army's definition of product revenue: directly selling a solution to solve a problem, not selling consulting, services, or research activities.
Other People's Money: Unsatisfactory "fails to present non-DoD sources for future R&D funding." Superior is "diverse and robust non-DoD sources of future R&D funding."
Taken together, 45 percent of your score sits in Programmatic Potential plus Commercial Potential. For a research-driven STTR team, that is the warning label on this topic. Nearly half the evaluation is not about the science. It is about whether you have done customer discovery with the Army, whether you have a named next deal, whether you have real product revenue history, and whether you have non-defense money in the picture. Academic strength does not compensate for a thin commercialization case, and the rubric is explicit that consulting and research revenue do not count as product revenue.
Contract Requirements After Award
Deliverables. Hardware or prototype deliverables per the topic, if applicable, with possible unique item identification requirements under DFARS 252.211-7003. Data deliverables can take the form of test data, computer software, algorithms, design details, progress reports, technical data, financial or management reports, or any information required by contract to be delivered. Data is ordered using DD Form 1423 Contract Data Requirements Lists (CDRLs) located at Section J, Exhibits of the contract. Each CDRL references a Data Item Description viewable through the Acquisition Streamlining and Standardization Information System (ASSIST).
All Army SBIR/STTR funding agreements include two CDRL requirements: Status Reports under DID number DI-MGMT-80368A, due at a specified time after award and periodically thereafter (monthly, bi-monthly, or quarterly), and a Scientific and Technical Report under DID number DI-MISC-80711A, a final report delivered per the prepared DD Form 1423. The Army end user may require additional deliverables including software documentation and user manuals, engineering drawings, operation and maintenance documentation, safety hazard analysis, or updated commercialization results.
Note that computer software and algorithms are named as deliverable data types. For a software topic with a university partner, that makes the data rights and allocation-of-rights conversation a contract-shaping issue, not a formality.
Meetings. A Start of Work Meeting within 30 calendar days of award, held at your facility unless otherwise designated, to establish mutual understanding of contract terms, conditions, line items, technical requirements, and sequence of events. Then periodic review meetings to report status, accomplishments, difficulties, and progress against planned goals and resources expended.
Invention reporting. Under FAR 52.227-11 and DFARS 252.227-7039, Phase II contracts require an Interim Report of Inventions and Subcontracts on DD Form 882 twelve months from the date of initial contract award. All Phase I and Phase II contracts require a Final Report of Inventions and Subcontracts on DD Form 882 within six months after completion of the contracted work. STTR awardees must report inventions within two months of the inventor's report to the awardee, and may do so through the Edison Invention Reporting System at iedison.gov. With a research institution partner, make sure both organizations know who is responsible for capturing inventor disclosures.
Controlled Unclassified Information. Successful firms must comply with CUI DoDI 5200.48. Firms must monitor CUI for aggregation and compilation based on the potential to generate classified information, report the potential classification of aggregated or compiled CUI to the ASA(ALT) Security Manager, and submit unclassified DoW information for review and approval for release in accordance with DoDI 5230.09. CUI records must follow approved mandatory disposition authorities. On a university-partnered project, CUI handling and publication review deserve an explicit plan, since the default academic instinct to publish collides with these obligations.
CMMC. The projected CMMC level requirement for this topic is Level 1. CMMC compliance support is one of the commonly requested TABA services, which is worth knowing if you are not yet compliant.
Asking Questions the Right Way
There are two channels and they have different rules.
Direct contact with topic authors, pre-release only. During pre-release, topic author names, phone numbers, and email addresses are published with the topic on DSIP. You may contact them by phone or email to ask technical questions about specific solicitation topics. Questions must be limited to specific information related to understanding a particular topic's requirements. You may not ask for advice or guidance on solution approach and may not submit additional material to the topic author. If information from an exchange is deemed necessary for proposal preparation, it will be made available to all parties through DSIP Topic Q&A. After pre-release, no further direct contact is allowed unless the author is responding to a question submitted during pre-release.
DSIP Topic Q&A. Opens on the pre-release date and closes two weeks prior to the topic close date, meaning October 7, 2026. Written questions and answers are posted on a non-attribution basis for public viewing. Answers are generally posted within seven business days and emailed to the inquirer. Questions are limited to technical information focused on understanding the topic's requirements. Requests for advice on solution approach, or administrative questions about SBIR or STTR program eligibility, proposal or cost structure, page count, budget and duration limitations, or the proposal due date, will not receive a response. Administrative questions go to the component-specific contact instead.
Once the proposal submission closes, no communication of any kind is allowed with the topic author or through Topic Q&A regarding submitted proposals.
The Army advises monitoring DSIP frequently throughout the solicitation period for updates and amendments to topics and for new Q&A entries.
Frequently Asked Questions
What is Army STTR topic ARM26TX06-NV003?
ARM26TX06-NV003 is an Army STTR Phase I topic titled "Agentic-AI, Schema-Driven Decision Management for Auditable Studies and Acquisition Decisions," released under the Department of the Army 2026 STTR Program Commercial Solutions Opening, Release 6. It seeks a demonstrated decision-as-a-program capability that converts complex engineering and acquisition questions into structured, auditable, signer-ready Decision Packages.
How much funding is available under ARM26TX06-NV003?
Phase I awards are up to $300,000 with a period of performance of 1 to 6 months. Up to $6,500 in Technical and Business Assistance funding may be requested in addition to the award ceiling. Proposals exceeding the topic's stated funding and duration limits will be deemed unresponsive. The $300,000 covers the combined effort of the small business and its research institution partner.
When is the proposal deadline for ARM26TX06-NV003?
Complete proposals must be certified and submitted in DSIP no later than 12:00 p.m. ET on October 21, 2026. That is noon Eastern, not end of day, which is 11:00 a.m. Central, 10:00 a.m. Mountain, and 9:00 a.m. Pacific. Proposals cannot be submitted after the deadline and will not be accepted or evaluated.
When does this topic open for proposals?
The topic opens for proposal submission on September 23, 2026, giving a 29 day submission window that closes October 21, 2026.
When does the DSIP Topic Q&A window close?
October 7, 2026, two weeks prior to the topic close date. Since answers are generally posted within seven business days, submit questions well before the cutoff if you need the answer in time to use it.
What is the difference between this STTR topic and the Army SBIR topic ARM26BX06-NV012?
The technical scope, funding ceiling, period of performance, and Phase I evaluation rubric are the same. The difference is the program. STTR requires a formal partnership with a research institution and a statutory division of work between the two organizations, and it allows the principal investigator to be primarily employed by either the small business or the research institution. SBIR has no partnership requirement and requires the principal investigator to be primarily employed by the small business.
Do I need a university partner to apply for ARM26TX06-NV003?
Yes. The STTR program requires a partnership with a qualifying research institution, which can be a university, a federally funded research and development center, or a qualifying nonprofit research organization. If you do not have a research institution partner, apply to the SBIR companion topic ARM26BX06-NV012 instead.
What is the required work split between the small business and the research institution?
Under the STTR program, the small business must perform at least 40 percent of the work and the research institution at least 30 percent, as a percentage of total effort. The governing definitions and requirements are in the DoW 2026 STTR Program CSO and the SBA SBIR/STTR Policy Directive rather than in the Army component-specific instructions, so verify the current terms there.
Can the principal investigator work at the university rather than my company?
Yes. The Army instructions state that primary employment of the Principal Investigator or Project Manager must be with the small business concern, or with the Research Institution in the case of STTR, at the time of award and during the conduct of the project. This is one of the meaningful advantages of the STTR path.
Are venture-backed companies eligible for this STTR topic?
Verify before you commit. The Army SBIR instructions for Release 6 include a provision confirming that firms majority-owned in part by multiple venture capital operating companies, hedge funds, or private equity funds are eligible for SBIR awards, subject to a joint venture eligibility certification. The Army STTR instructions for the same release do not contain that provision, and the majority venture-capital ownership authority is SBIR-specific under the statute. Check your eligibility against 13 CFR 121.702 and the SBA SBIR/STTR Policy Directive.
Is Direct to Phase II available for ARM26TX06-NV003?
No. The Release 6 topic index lists this topic as Phase I (Feasibility). Direct to Phase II proposals may only be submitted where a topic posting explicitly allows them. Phase II proposals may otherwise only be submitted by Phase I awardees, and unsolicited Phase II proposals will be rejected without evaluation.
How long is the technical volume?
Seven pages for a Phase I proposal, including all required sections. Appendices and supplemental content count toward the seven pages. Cover sheets and tables of contents may be excluded at the Government's discretion. Minimum 10-point type, 8.5 by 11 inch pages, one-inch margins including the header.
What are the required sections of the technical volume?
In this order: Introduction, Army Benefits, Technical Approach, Programmatic Potential, and Commercial Potential. These map directly onto the Phase I evaluation rubric in Appendix A of the Army instructions.
Where do I submit my proposal?
Through the Defense SBIR/STTR Innovation Portal (DSIP) at dodsbirsttr.mil. Proposals submitted by any other means will be disregarded. Select Army as your component and ASA(ALT) as your command.
What registrations do I need before submitting?
An active SAM registration in accordance with FAR Provision 52.204-7, and individual DSIP user accounts for everyone preparing or submitting the proposal. Your DSIP account profile must match your SAM record, including CAGE code, UEI, legal business name, DBA name, and physical address. Register early, since mismatches can delay or void an award.
What are the most common ways proposals get disqualified?
Missing a mandatory Volume 5 attachment, since a completed DSIP submission does not mean supporting documents were uploaded. Failing the technical volume formatting requirements. Exceeding the funding or duration limits. Leaving the proposal in "In Progress" or "Ready to Certify" status instead of certifying it. Incomplete or inaccurate Volume 7 foreign affiliation disclosure. Failing to identify foreign nationals or dual citizens on the project, including research institution personnel.
Do I have to disclose foreign nationals at my university partner?
Yes. The List of Foreign Citizens requirement names Research Institutions explicitly, and the Volume 7 covered-individual definition applies to anyone contributing substantively to the project whether employed by the prime, a subcontractor, a consultant, or a partnering research institution. Request a complete roster with citizenship and visa status from your partner early.
How is the Phase I proposal scored?
By weighted rubric: Technical Approach 35 percent, Commercial Potential 25 percent, Programmatic Potential 20 percent, Army Benefits 15 percent, Introduction 3 percent, and Proposal Quality 2 percent. Each element is rated Unsatisfactory, Marginal, Satisfactory, or Superior.
Do I need to talk to Army stakeholders before I submit?
Effectively yes. Army Customer Discovery and Validation sits inside the 20 percent Programmatic Potential weight, and the rubric scores "no customer interviews completed, no validation" as Unsatisfactory while reserving Satisfactory and Superior for extensive or exhaustive interviews with validation informing a credible transition strategy.
Can I contact the topic author?
Only during the pre-release period, meaning before the topic opens on September 23, 2026, when author contact details are published with the topic on DSIP. Questions must be limited to understanding the topic's requirements. You may not ask for advice on your solution approach or send them material. After pre-release, use DSIP Topic Q&A, which closes October 7, 2026.
When will I hear whether I was selected?
Within 90 days of the October 21, 2026 close date, so on or about January 19, 2027. Notification goes by email to the Corporate Official listed on your proposal cover sheet. A selection notice is not a contract and is not authorization to commit or expend funds.
Will I get a debrief if I am not selected?
You will get feedback, not a debrief. The Army provides feedback per the SBA SBIR/STTR Policy Directive, with instructions in the selection decision notice, and offerors are entitled to no more than one feedback per proposal. Competitive procedures here are governed by the SBIR/STTR Policy Directive, so FAR Part 15 debriefings are not available.
What kind of contract will I receive?
A FAR-based, firm-fixed-price contract, with payments tied to measurable milestones or deliverables agreed to with the Government. The Contracting Officer retains the right to negotiate contract type and price.
What is TABA and how do I get it?
Technical and Business Assistance is funding provided in addition to the award ceiling, up to $6,500 per Phase I project, for services such as market research, IP support, CMMC compliance, DCAA compliance, and marketing. You can use an Army preferred vendor, in which case the cost stays out of your Cost Volume, or your own provider, in which case the cost must be in Volume 3 and a two-page rationale must be in Volume 5. TABA is denied if the costs are omitted from the initial submission or exceed the cap.
What happens after Phase II?
The Army anticipates follow-on Phase III contracts for individual customer installation, comparable to how large ERP deployments from SAP and Oracle are rolled out. Separately, the Army Phase II Enhancement Policy may provide up to $500,000 in matching funding on a dollar-for-dollar basis against committed non-SBIR/STTR funding from a DoW component, another federal agency, or a commercial investor.
What CMMC level does this topic require?
The projected requirement is CMMC Level 1. CMMC compliance assistance is one of the frequently requested TABA services if you need help getting there.
Are there prohibited technologies I need to avoid?
Yes. Proposers and their subcontractors are strictly prohibited from using products or services listed in RFO FAR 40.202, including but not limited to TikTok/ByteDance, Kaspersky, and specific unmanned aircraft systems. Offerors using prohibited sources will be eliminated.
Who do I contact with administrative questions?
Email usarmy.SBIRSTTR@army.mil for questions about Army STTR Program administration and the Army component-specific proposal instructions, no later than 15 days prior to solicitation closing, which is October 6, 2026. For DSIP portal problems, email DoDSBIRSupport@reisystems.com during normal operating hours, Monday through Friday, 9:00 a.m. to 5:00 p.m. ET. The Army SBIR/STTR Office is at 2530 Crystal Drive, Suite 11192, Arlington, Virginia 22202.
Positioning Advice for Startups Considering This Topic
Everything that applies to the SBIR twin applies here, plus a set of considerations specific to running a two-organization proposal.
Decide honestly which program fits. The STTR partnership requirement is a feature if your research institution brings people, instruments, datasets, or methods you cannot replicate, and a liability if it is decoration. Evaluators read Technical Team and Technical Approach closely, and a partner with no defined role is visible.
Make the partnership do work in the narrative. Name what the research institution contributes that you cannot: the semantic modeling expertise, the knowledge graph work, the evaluation methodology, access to a laboratory or a corpus. Then show the seam, meaning how their contribution integrates with what you build. A proposal that reads as two projects stapled together scores worse than a single-organization proposal.
Start the institutional clock immediately. Sponsored programs offices have their own internal deadlines, often a week or more before the sponsor deadline, and they control whether your budget and any required agreements arrive on time. Call them before you write anything.
Resolve intellectual property early. Phase II envisions a commercial open-source platform, Phase III envisions installation contracts, and the deliverables include computer software and algorithms. Your partner's technology transfer office will have views on all of that. Better to hear them in September than during contract negotiation.
Lead with reproducibility and auditability, not model capability. The word "auditable" is in the topic title. "Reproducible outputs," "traceability," "human control," and "predictable behavior" all appear in the description. The differentiator is not a better model, it is a governed system where a decision authority can see why the answer is the answer and what would change it.
Build the "what flips the decision" logic explicitly. This phrase appears verbatim in the Phase I desired outcomes. If your proposal does not have a named capability that answers it, you are leaving a scoring element on the table.
Plan for two demonstrations, and make them structurally different. One point-in-time trade study, one multi-episode decision program with a refresh cycle. Naming both, with distinct data and distinct evaluation runs, directly satisfies a stated Phase I outcome.
Speak ground vehicle. The topic calls for context engineering specifically for the ground vehicle domain through knowledge graphs and semantic layers. Generic decision-support framing will read as misaligned, and Alignment is scored.
Ground yourself in Digital Engineering standards. SysML 2.0 and UML are named. Digital threads and digital twins are named. Interoperability across DE tools is an explicit expectation.
Use proven, ideally Army-fielded, enabling technologies. The rubric penalizes emerging or cutting-edge dependencies and rewards Army-fielded ones. This is where academic instincts can cut against you, since novelty is the currency of research and proven maturity is the currency of this rubric. Frame the novelty in the schema and the governance, not in the underlying infrastructure.
Do not let the research strength mask a weak commercialization case. Commercial Potential is 25 percent, and the Army defines product revenue as directly selling a solution, explicitly excluding consulting, services, and research activities. Sponsored research funding is not product revenue and grant funding is not "other people's money" in the sense the rubric means. The automotive vehicle design analog in the Phase III dual-use paragraph is handed to you. Use it.
Do the customer discovery now. Programmatic Potential is 20 percent and the top of the scale requires exhaustive interviews with validation that informs a credible transition strategy. PAE Maneuver Ground and PAE Fires are named in the Phase II text, which tells you exactly where the Phase II cases will come from and therefore who is worth talking to during pre-release.
Position against, or alongside, DAOSoft. The Army names the current COTS baseline and states it is agnostic and open to any and all ideas that extend or replace current acquisition capabilities. That is an invitation to make a clear-eyed competitive argument, including against indirect substitutes, which the Solution's Advantages criterion explicitly asks about.
