Portfolio
Most of this work lives in private repositories serving real workloads — so each flagship below gets a full case study: the problem, the architecture, and what actually shipped. Several have interactive demos.
Opportunities in seconds. Pink Team drafts in minutes.
Finds federal opportunities the moment you ask, profiles vendors and competitors, tracks bid protests, and drafts Pink Team proposal content and capture briefs backed by 28.6M contract awards. 90+ research tools inside ChatGPT and Claude.
28.6M
Contract awards indexed
90+
Research tools exposed
Accurate answers from your documents, automatically
A smart search engine for your company documents that tests and improves its own answers. Built for industries where a wrong answer is expensive: healthcare, finance, and government.
18
Architecture decision records
10
Pipeline stages instrumented
Evidence-based answers with citations you can check
Evidence-based research for healthcare, science, and technology questions. Searches six scholarly databases at once (PubMed, arXiv, and more), scores the quality of its own answers, retries until they meet the bar, and cites exactly where every claim came from.
15
MCP tools
13
Scholarly + technical sources
Know which AI to trust before you bet on it
Four leading AIs predict IPO outcomes, then get scored against what actually happened. 800+ IPOs backtested, running live on a public dashboard anyone can check.
800+
IPOs in research universe
4
LLM arena models
Pass AI audits with evidence, not promises
Everything you need to prove your AI is under control: risk assessments, policies, and training records mapped to federal requirements. Ready when the auditors ask.
8
Governance MCP tools
6
Framework crosswalks
Days of research done in hours, every answer checked
Six AI specialists research biotech questions together, and a referee AI checks every answer before you see it. Built for research teams that cannot afford confident nonsense.
6
Specialist agents
5
Judge evaluation axes
Also built
Shared data platform: 47 ingestion pipelines, a 15-step autonomous daily refresh with quality assertions and anomaly alerts, and 39 service modules feeding four downstream AI products from one PostgreSQL schema.
47
Ingestion pipelines
30M+
Rows under management
Design for a data-as-a-service MCP gateway federating ~280 federal, legal, corporate, scholarly, and OSINT feeds into one endpoint with an entity-relationship graph, attachment pipeline, and persistent memory — ~145 tools across 12 namespaces.
~280
Federated sources (design)
~145
Tool surface
MCP servers for Microsoft Graph, SharePoint, and Power BI: an AI agent can search SharePoint, move files, create push datasets, and assemble a live Power BI dashboard from a single natural-language request.
15
Power BI tools
Entra ID
Auth
LEAPS options tracking system with OAuth-authenticated live broker quotes, two years of price history, 19 daily technical indicators across 2,630 contracts, and a 21-second automated daily pipeline feeding a five-tab dashboard.
2,630
LEAPS contracts tracked
21s
Daily pipeline runtime
Enterprise RAG system ingesting GitHub repositories and documents with markdown- and code-aware chunking, pluggable embeddings (Bedrock Titan or local), and cited Q&A over institutional knowledge via FastAPI.
Pluggable
Connector pattern
2
Embedding providers
Reusable scaffold for document-intelligence products: hybrid pgvector + BM25 retrieval with score fusion, Neo4j knowledge graph, FastAPI + React, and a config-only AWS migration path. Spawned the RFP-intelligence product line.
3
Retrieval indexes
docker compose up
Spin-up
Five interactive demos run right in your browser — the same workflows, tool-call patterns, and quality gates as the production systems.
Open the product demos →