Belleau Labs
Back to contracts

Making Petabytes of Data Useful in Near-Real Time

AFWERX · AFWERX STRATFI · AFWERX

AI-Readiness Score
17/25
Pathway Speed
5/5
Timeline Realism
3/5
Problem Framing
3/5
AI / ML Fit
3/5
Award + Transition
3/5

Award

$1,247,391
Award ceiling
PROJECT EIGHT NINETEEN, LLC
Awardee
Posted May 20, 2025

Description

Project Eight Nineteen proposes Data Prism, an AI-driven data integration and decision-support platform designed to transform petabytes of fragmented, unstructured ISR, cyber, and operational data into actionable intelligence in near-real time. Built on MinusOneDB (m1db), Data Prism eliminates the performance bottlenecks of traditional data warehouses by leveraging a search-based primary data store, high-throughput indexing, and federated storage. This ensures continuous data ingestion, normalization, and AI/ML readiness, even in denied, degraded, intermittent, and low-bandwidth (DDIL) environments. AFSOC’s intelligence cycles are currently 40% slower due to reliance on siloed, manually processed data, hindering decision-making and reducing mission effectiveness. Data Prism automates data fusion, eliminating 80% of manual processing and enabling autonomous AI-powered intelligence at the tactical edge. This capability is critical as near-peer adversaries, such as China, operationalize AI-driven warfare with real-time decision engines. Phase II will develop, test, and deploy Data Prism within AFSOC and the Chief Digital and AI Office (CDAO), ensuring alignment with JADC2 and DoD AI integration initiatives. With a scalable, containerized microservices architecture, Data Prism supports multi-domain intelligence fusion, enhancing interoperability across joint, allied, and coalition forces. The platform will transition into a Program of Record via TACFI and STRATFI funding, ensuring long-term defense integration while driving commercialization in finance, healthcare, and logistics.

Score Rationale

STRATFI is a top-tier fast-track instrument, earning the highest pathway score, and the explicit mention of TACFI/STRATFI transition into a Program of Record with CDAO and AFSOC end-users named is a meaningful transition signal. However, the problem framing mixes genuine AI/ML work (data fusion, autonomous intelligence at the edge) with heavy data engineering and infrastructure work — MinusOneDB and federated storage are real but the '80% manual processing reduction' metric is unverified marketing language, not a bounded success criterion. The award ceiling of ~$1.25M is modest for the scope claimed (petabyte-scale, DDIL, multi-domain fusion), and the unknown response deadline prevents a confident timeline realism score above mid-range.

Source

View original posting
Back to all contracts