How I Built UseDocs on Cloudflare: RAG, Agents, and Multi-Tenant Edge
The real architecture behind UseDocs: crawl → embed → cited answers on Cloudflare, with Durable Objects, Vectorize, metering, and honest escalation when confidence is low.

Founding Full Stack Engineer
Zero-to-One Products · Fintech · SaaS · AI
Founding engineer with 7+ years building products from 0→1 across fintech, blockchain, SaaS, AI, and workflow automation. Repeated day-one / early-team ownership: architecture, shipping, auth, payments, APIs, and production infrastructure — not just tickets.
Comfortable wearing product + eng hats, iterating fast with founders, and owning systems end-to-end from MVP through scale. Startup strengths: zero-to-one delivery · full-stack ownership · product judgment · ambiguous problem-solving · shipping under constraints.
Outside of code, I recharge by traveling, exploring nature, and spending time with my kid—my best source of inspiration.
The real architecture behind UseDocs: crawl → embed → cited answers on Cloudflare, with Durable Objects, Vectorize, metering, and honest escalation when confidence is low.
How dynamic routes work on Cloudflare Workers and Pages — path params, catch-alls, custom domains, and multi-tenant routing without drowning in config.
Reactivate (APEX-TEK)
Blockpeer Finance
Stealth Startup (Fintech / Blockchain)
IVP
AI support platform (RAG): multi-tenant crawl → cited answers → honest escalation when low confidence. Cloudflare architecture with usage metering, embeddable widget, Slack/Discord/Teams, and a docs-ops loop for gaps, digests, and analytics.
Cloudflare Workers, Agents SDK, Durable Objects, Workers AI, Vectorize, R2, D1, Queues, React, TypeScript, Vite, Better Auth, Dodo Payments
AI web product with Supabase workflows and Gemini-powered production UX.
Next.js, Supabase, Google Gemini AI
AI image product with API generation pipelines and responsive UI on Vercel.
Next.js, Python, Tailwind CSS, DaisyUI, Vercel, Stable Diffusion API
Sharda University
2014 – 2018