The Engineering Architect
Designs agentic LLM systems tuned for real-time UX — multi-modal, tool-using, latency-obsessed. Shipping production AI since before RAG had a name.
- Agents
- Latency
The Order of Treasure Scouts
You’ve heard of the work. Not the names.
We’ve founded and sold companies, run growth at a B2C unicorn, and shipped production LLM and RL systems since the GPT-2 era. The companies you’d recognise; we’d rather you judge this one on the board.
4
Operators
3
Exits between us
2019
Shipping AI since
0
Names on this page
The Operator
Ran growth — engineering and product — at a B2C unicorn. Doubled paid-marketing efficiency in a quarter. Knows where the leverage in a funnel actually sits.
- Growth
- Funnels
The Founder
Built a B2B web-intelligence company acquired by a tech unicorn. Raised from top-tier funds. Has been in the room at every stage from whiteboard to wire.
- Exited
- Web data
The AI Researcher
Used RL to train the best model on the planet for ████████████. Builds models that learn from their own mistakes, not just other people’s writing.
- RL
- Evals
Some of us wear more than one of these hats. Some of these hats wear more than one of us. You’ll figure it out.
A note for investors
We are not raising.
TreasureScout is one of a handful of things this band is quietly compounding. Price intelligence as a consumer product is, in our read, deeply undervalued — but we’re builders first and fundraisers second, and we like to earn the right to the conversation.
If you’re the kind of investor who likes to be early, quiet, and right — write anyway. Say “investor” in your note and the Scout routes it to the right hands.
The Scout will route it