The Problem We're Solving

Existing AI tools treat research as question-answering. For serious researchers, journalists, and knowledge workers, that's not enough — citations get hallucinated, sources go unverified, and there's no real "human in the loop." Ubik Studio is built around the thesis that AI should assist rigorous research workflows, not replace judgment.

My Role

As the sole technical co-founder, I own the entire technology layer: from architecture decisions down to production deploys.

• Designed and built the full-stack application from scratch (Next.js frontend, Python backend)
• Set up and maintain all infrastructure on Cloudflare Workers and edge compute
• Built the multi-model routing system that lets researchers choose between providers
• Created the PDF annotation pipeline and citation verification engine
• Made every architectural call — database choice, deployment strategy, API design

[TODO: What does a typical working week look like now that the product is live? How does your time split between product, engineering, and other co-founder responsibilities?]

Key Technical Decisions

[TODO: Walk through 2–3 of the most important architectural choices you made and why — e.g. local-first vs cloud-first, choice of LLM routing approach, why Cloudflare Workers specifically]

What I'm Most Proud Of

[TODO: Pick 1–2 things about this product or your engineering work that you'd point a strong engineer to and say "this is the hard part we solved well"]

Traction & Status

[TODO: Is Ubik Studio live in production? How many users / research teams are using it? Any paying customers, institutional pilots, or press coverage worth mentioning?]