Research systems
Find the signal.
Keep the evidence.
Research gets more useful when every accepted fact has a source, a timestamp, and a reason to trust it. That is the standard I designed into Bloodhound.
Bloodhound
Bloodhound is an autonomous research and lead-intelligence system documented in my project archive. It plans missions before searching, crawls public sources on its own infrastructure, stores accepted facts with evidence, resumes after interruption, and exports durable results.
Public sourcesLead researchEvidence-backedFrom a question to an inspectable result
The system normalizes results from approved search sources, deduplicates companies and contacts, records confidence, and produces JSON, CSV, and Excel exports. A private shared knowledge service reuses only fresh, high-confidence records.
Its crawler respects robots.txt, rejects private network addresses, limits scope, retries temporary failures, extracts source text and JSON-LD, and keeps a source manifest. The point is not to make research look magical. It is to make the reasoning trail inspectable.
Safety boundaryWhat it does not do
The documented system does not bypass authentication, CAPTCHAs, paywalls, robots rules, or access controls. Qualified leads require public evidence and a confidence threshold. Those constraints are part of the product, not an afterthought.
Where this helps
Founder research, market mapping, lead discovery, public-source due diligence, and any workflow where “here are some names” is not enough. I care about the source trail because people need to decide what to do with the answer.