Startups stop pitching investors who were never a fit.
- Before
- Startups pitched investors who were never a fit.
- I built
- I built a pipeline that researches every VC fund.
- Result
- 5,000 VC funds, ranked for each startup, every time.
Systems I've built for clients.
Each one starts with research.
ResearchTech Credora Credora enters a stablecoin name. Their spreadsheet fills itself.
- Before
- Their team filled every field by hand.
- I built
- A pipeline that researches every field on their list.
- Result
- Every value comes with its sources and a confidence level.
Hacken I found 400+ spots where a blockchain engine gives different results.
- Before
- A blockchain engine must give the same answer on every run.
- I built
- Agents guess where it might give a different answer on a re-run.
- Result
- 400+ spots found and confirmed.
Other AI systems
Email handling, chatbots, testing agents, video analysis and more.
Email handling in a CRM
~2,000 emails a day, sorted. Support replies in seconds instead of hours.
Healthcare chatbot
Patients book a visit in one conversation. It hands over to a real person when needed.
Email order intake
A 2-hour daily job now takes ~2 minutes. Matches a manufacturer's email orders to its products, and drafts replies.
AI that tests software
Clicks through apps like a person, fills in forms and checks them.
Game feedback from video at scale
Analyses every frame of gameplay videos on YouTube and Twitch. Finds frustration, bugs and early feedback for game developers.
Invoice extraction
Reads invoices and puts the data into tables.
More under NDA. Ask on a call.
Need a system like these?
Teams I've trained
Workshops for engineering teams.How to code with AI tools, set up a harness, and do much more work with much less effort.
