Full-Stack & AI Engineer · Founder, Synksha
Delhi, India · 8 years building production systems · Open to remote founding-engineer roles
I design, build, and ship AI products end-to-end. Most recently I built Synksha — a production AI consumer platform — solo, from architecture to launch.
A live AI platform (on the Google Play Store, payments integrated). Sole engineer across product, infrastructure, and ML systems.
Stack: Next.js 15 · TypeScript · Tailwind · Prisma + PostgreSQL (~79-model schema) · GCP Cloud Run · PostHog
LLM systems: a three-tier model-routing layer over the Anthropic Claude API (Haiku / Sonnet / Opus) that picks the right model per task to balance cost, latency, and quality
Memory: a 10-layer memory architecture giving the product durable, personalized context across sessions, with cached layer summaries for token control
Safety: a crisis-detection layer that short-circuits before any model call and routes at-risk users to support — safety decoupled from model output
Integrations: Razorpay payments · WhatsApp Business API
8 years full-stack across product startups and scaled teams — including FanCraze (ICC cricket marketplace on Flow blockchain, where I built and scaled the most-used features) and payment-gateway / serverless work that moved real revenue and cost numbers.
TypeScript · Node.js · Next.js / React · Prisma / PostgreSQL · LLM application engineering · multi-model routing · memory systems · GCP · AWS · Docker
Currently exploring founding-engineer roles at remote-first, AI-native startups. If you're building one, let's talk.
