Back to home pageBack to home page
Pick one to see where and how I've done this work.
- PrizervFounding AI engineer and sole builder on Prizerv, a four-person venture building an AI self-discovery product: built the whole product end to end with AI tools. Three guided AI conversations run on Next.js and Postgres with row-level security, and a server-side chain extracts signals from each one, scores coherence and streams back a written report.
- Slurrp FarmFor Slurrp Farm, a $90M consumer foods brand, found the gap, customer signal scattered across platforms and read reactively if at all, then built the answer with AI end to end. Scrapers and ingest scripts feed a 15,000-plus verbatim corpus in Postgres, and an agent answers product questions from it with full-text retrieval, evidence counted in code and a writing model that sees only that frozen evidence pack.
- Loop HQAn autonomous AI agent workflow I built that now runs the coding work across six products: builder and independent reviewer agents working for hours unattended, and nothing reaching production without a person's decision.
- Slurrp FarmWorked directly with the founders and senior leadership of Slurrp Farm, a $90M consumer foods brand, to scope the problem, then worked with stakeholders across departments to pull their customer data together and built two AI tools on it, Mill and Kernel.
- Share Our Strength (India)Pitched, built and demoed a donor intelligence platform for Share Our Strength (India), a fundraising nonprofit: a Monday briefing for its Executive Director with a ranked reach-out list and a drafted opener for each name. Every fact shows its source and date, and relationships are rated "likely knows" or "possible overlap".
- LeverBuilt Lever, a tool that answers the question that comes before build versus buy, what AI can do for a business at all, with each opportunity ranked by fit, likely return and what would sink it.
- Business school AI strategyProposed AI-powered tools and workflows to a business school's careers team as an AI strategy: three moves in a deliberate order (an advisor cockpit, then employer and alumni contacts, then targeted role lists), each earning the next, plus a list of what to skip.
- Slurrp FarmWorked directly with the founders of Slurrp Farm, a $90M consumer foods brand, to understand where and how the business was using AI, if at all, identified where AI could add the most value, then designed and built its first ever AI products.
- My notes on AI strategyTwo LinkedIn notes, on where AI adoption gets held up and on the three AI windows.
- PrizervBuilt Prizerv, an AI self-discovery product in three phases, Being, Doing and Becoming, where each adaptive AI conversation carries the earlier phases' profiles forward. Reports are written from signals extracted once a session closes, and a quote is shown as the user's own words only if code finds those exact words in what they typed.
- Slurrp FarmBuilt Mill and Kernel for Slurrp Farm, a $90M consumer foods brand. Mill merges scraped reviews, call transcripts, surveys and interviews into a 15,000-plus verbatim corpus with every claim traced to its sentence; Kernel tests a product idea against that corpus, with code counting the evidence, a model writing only from that frozen pack, and both a fixed pipeline and an agentic path that chooses its own steps.
- HopBuilt Hop, a London night-out app for a group: three answers and no account, five picks a week that each say why they fit, and one link that checks everyone's limits before it reaches the group chat, which friends can answer without the app.
- Built for myselfBuilt two apps for my own daily use: Fuel for food and training, and Tally for shared money.