Ayush Sharma

About Ayush Sharma

Full-Stack Developer and AI Engineer with a backend and systems focus. Currently CTO at Kakiyo OÜ and creator of memcontext.in, a persistent memory layer for AI coding agents.I build scalable, backend-heavy AI systems engineered for real-world scale, low latency, and solid infrastructure. Clean APIs, sane databases, cache layers that work, queues that don\'t lose data, and a frontend that sits cleanly on top. I care about applied AI, end-to-end execution, and shipping to production, not demos.Impact at Kakiyo OÜ (CTO):• Cut P50 latency from 2.8s to 20ms, P90 to 90ms• Migrated millions of rows from Appwrite to PlanetScale with zero downtime• Reduced infrastructure cost by 50%• Built the hiring pipeline and lead engineering end to endThings I\'ve built and shipped:• MemContext, persistent memory for AI coding agents. Lets agents remember context across sessions via vector search and MCP integrations.(Hono, Next.js, Drizzle, Postgres)• Zenox, agent orchestration plugin for OpenCode with specialized subagents and background tasks.(TypeScript, MCP, Bun)• Plnr, AI planning and security CLI that turns codebase context into concrete implementation plans.(Node.js, OpenRouter, Exa AI)• CappyChat, production AI chat workspace with multi-model routing, realtime sync, and tool calling.(Next.js, Appwrite, Upstash Redis)• Bucket Buddy, S3-style storage manager for AWS and Cloudflare R2.(Next.js, Prisma, Postgres, AWS SDK)Backend & systems: Node.js, Hono, Next.js API routes, Postgres, MongoDB, Prisma, Drizzle, Redis, Docker, AWS, PlanetScale, GraphQL, system design, API design, caching, database migrations, observability.AI engineering: RAG, vector search, agent orchestration, multi-model routing, MCP servers, Vercel AI SDK, OpenRouter, LangChain, OpenAI, agentic workflows.Frontend: Next.js, React, TypeScript, Tailwind, Zustand.I also write about systems I find interesting. Recent posts cover how WebRTC actually works (signaling, ICE, SFU vs MCU) and webhooks done right (transactional outbox, retries, idempotency).Open to full-stack, backend, and AI engineering roles, freelance, and advisory work. Based in Jaipur, India.Portfolio: aysh.me

Experience

1. Chief Technology Officer

   Kakiyo

   Apr 2026 — Present · San Francisco, CA, US

Education

• Manipal University Jaipur

  Bachelor of Technology - BTech, Computer Science & Engineering (IOT & IS)

  2023 — 2027