Claude agent mines USAspending federal contract data for automated lead gen
JanJanJaJa · reddit · 2026-08-18
A subcontractor shared a Claude-based lead-gen agent pipeline whose key insight: the hard part is the data, not the agent.
- Data source: US federal award data is free via USAspending's JSON API (from FPDS); Claude wrote spendingbyaward queries in 20 minutes. SAM.gov's API is tightly rate-limited (10 req/day public key).
- Key filter: searching "data center" returns hundreds of useless IT-services contracts; crossing with construction NAICS codes (236220, 238210) yields a readable shortlist (e.g., Trevino Group $48.9M, Hurley JV $27.8M).
- Surprise find: 7 of 12 filtered awards in one window were VA, 5 tied to the EHRM program — a predictable, recurring source of data-center construction work.
- Contacts: award records only list the contracting officer; browser use fills the gap by finding subcontract/supplier contacts on winners' sites, with mediocre hit rates. For anything with an API, browser automation is the wrong tool — the page-scraping v1 broke on first layout change, the API version never did.
- Email: the agent sends and reads replies via its own Atomic Mail inbox, later moved to the company domain for credibility.
More from coding & agent
- Harness Co-Training: Models shaped inside agent loops become new industry norm — DynamicWebPaige · 2026-08-18
- Research Blog: How Native Memory Improves Agent Search Efficiency — realJessyLin · 2026-08-18
- ChatGPT Keeps Forgetting Plugin Instructions: A Frustrating Tool Loop — Angaisb_ · 2026-08-18
- QM v0.1.5 Released: Adds Multi-Pane Sessions and MCP Connectors — ycombinator · 2026-08-18
- AI-HQ routes tasks away from LLMs for deterministic efficiency — Dismal-Reveal-6898 · 2026-08-18
- When to use an AI Agent vs a Skill — ColbyHawker · 2026-08-18