Compositional Learning-And-Reasoning for AI Complex Systems Engineering (CLARA)

SOL #: DARPA-PA-25-07-02Solicitation

Overview

Buyer

DEPT OF DEFENSE
Defense Advanced Research Projects Agency (Darpa)
DEF ADVANCED RESEARCH PROJECTS AGCY
ARLINGTON, VA, 222032114, United States

Place of Performance

Place of performance not available

NAICS

Research and Development in the Physical (541715)

PSC

National Defense R&D Services; Department Of Defense Military; Basic Research (AC11)

Set Aside

No set aside specified

Timeline

1
Posted
Feb 10, 2026
2
Last Updated
Mar 11, 2026
3
Submission Deadline
Apr 17, 2026, 4:00 PM

Qualification Details

Fit reasons
  • NAICS alignment with historical contract wins in similar service areas.
  • Scope strongly matches core technical capabilities and delivery model.
Risks
  • Past performance thresholds may require one additional teaming partner.
  • Potential clarification needed on staffing minimums before bid/no-bid.
Next steps

Validate eligibility requirements, assign capture owner, and schedule partner outreach to confirm teaming strategy before submission planning.

Quick Summary

The Defense Advanced Research Projects Agency (DARPA) Defense Sciences Office (DSO) is seeking innovative basic or applied research concepts for Compositional Learning-And-Reasoning for AI Complex Systems Engineering (CLARA). This Disruption Opportunity (DO), issued under DARPA-PA-25-07, aims to develop high assurance artificial intelligence (AI) systems through a scientific, theory-driven architectural foundation for composing diverse Machine Learning (ML) and Automated Reasoning (AR) subsystems. Awards will be made as Other Transactions (OT) for Prototype projects, with a total value up to $2,000,000. Proposals are due April 17, 2026.

Scope of Work

The CLARA program focuses on two technical areas:

  • Technical Area 1 (TA1): Develop new approaches for high assurance ML/AR (AR-based ML), encompassing theory, algorithms, open-source software creation, and demonstrations. This includes ensuring models are adaptable and manually updatable.
  • Technical Area 2 (TA2): Develop a software composition library to support inter-approach compositions, integrating validated TA1 tools. This involves creating common structures, terminology, data formats, and interfaces. Performers are also expected to participate in program-wide activities such as kickoff meetings, workshops, and hackathons. Software developed is expected to be open-sourced.

Contract & Timeline

  • Type: Other Transaction (OT) for Prototype project
  • Total Award Value: Up to $2,000,000 (combined Phase 1 base and Phase 2 option)
    • Phase 1 (Feasibility Study): 15 months, up to $1,350,000
    • Phase 2 (Proof of Concept): 9 months, up to $850,000
    • Combined duration not to exceed 24 months.
  • Set-Aside: None specified (open to all responsible sources)
  • Proposal Due: April 17, 2026, 4:00 p.m. ET
  • Target Award Date: June 16, 2026
  • Anticipated Program Start: June 22, 2026
  • Published: March 11, 2026

Submission & Evaluation

Proposals must address either TA1 or TA2; organizations cannot be selected for both. An Independent Verification and Validation (IV&V) team will be competed separately. Abstracts are strongly encouraged, with a specific template provided. Cost proposals must utilize the provided Streamlined Cost Buildup Workbooks (TA1 and TA2 specific). Evaluation will be based on criteria outlined in Section 6 of DARPA-PA-25-07. An Associate Contractor Agreement clause emphasizes collaboration and information exchange among performers.

Eligibility

DARPA encourages solutions from all responsible sources. Specific instructions and updated eligibility criteria are provided for FFRDC/UARC/Government Entity proposers, who are generally discouraged unless by exception.

People

Points of Contact

BAA CoordinatorPRIMARY

Files

Files

Download
Download
Download
Download
Download
Download
Download
View

Versions

Version 2Viewing
Solicitation
Posted: Mar 11, 2026
Version 1
Solicitation
Posted: Feb 10, 2026
View
Compositional Learning-And-Reasoning for AI Complex Systems Engineering (CLARA) | GovScope