I’ve been interested in OSINT (open source intelligence) for years. Not professionally. I’ve just always been fascinated by how much public information exists, what it means for privacy and security, and the idea that regular people should have access to the same investigative tools as journalists, law firms, and government agencies.

I’d been trying to build something like Samaritan (that’s what I’m calling it) for a long time. An application that could simplify common OSINT tasks, run lookups across multiple sources, and package results into something usable.

Who is this OSINT tool for?

At first it was just for me. Then I started to see how useful it could be across industries: HR, legal, real estate, personal safety, political research, journalism.

I’ve spent years building a catalog of OSINT resources, specialty databases, and investigative tools. The catalog was the asset. I just needed a platform that could actually put it to work.

Then there was the vision problem. I grew up on sci-fi computers. I’ve played Watchdogs. I’ve seen every episode of Person of Interest. I believed a real intelligence platform you could actually operate was buildable with the right stack.

I tried GPTs, other automation platforms, and agentic tools, but nothing handled the actual workload in a way that made sense.

And finally, maybe most importantly: I’m obsessed with running my own AI models and agents on my own hardware. Keeping my investigations, my sources, and my work off someone else’s servers. In a world where every AI tool wants to be cloud-connected and subscription-gated, local AI isn’t just a technical preference. It’s a privacy stance.

When OpenClaw came along, I knew it was the platform I’d been waiting for to build this OSINT automation tool.