Tanmay Singewar

Full-stack engineer building AI agent systems and product infrastructure

HSR, Bengaluru India

  1. I build full-stack products and AI agent systems where runtime behavior, tools, workflows, and user experience need to hold together.
  2. Currently a Full-Stack & AI Agent Engineer at VegaStack, working on DevOps automation, platform systems, and internal AI tooling.
  3. My strongest work sits between product engineering, backend systems, and agents: auth, billing, HITL flows, memory, tool execution, deployment, and real-time dashboards.
  4. Open to full-time or contract roles around AI agents, developer tools, and scalable product engineering.
GitHubLinkedInResumeEmail

Work Experience

Sept 2025 - Present

Full-Stack & AI Agent EngineerVegaStack

HSR, Bengaluru

Building AI Agents for DevOps automation, platform systems, secure internal AI tooling, billing flows, and real-time operational dashboards.

  • Shipped agentic DevOps workflows with controlled credential access, middleware, and HITL debugging loops.
  • Worked across auth, billing, backend optimization, security middleware, and real-time operational dashboards.

Company projects

DevOps Deep Agent

Built AI-assisted DevOps tooling with custom middleware, controlled credential access, and HITL debugging workflows.

VegaStack Platform

Worked on auth, billing, caching, security middleware, and backend workflows across production systems.

Chaabi

Improved backend APIs, RBAC ticket management, and real-time analytics dashboards.

Next.js, TypeScript, PostgreSQL, Stripe, BetterAuth, TailwindCSS, Python, LangChain, Cloudflare

Sept 2024 - April 2025

Full Stack DeveloperCultureVo

Remote

Built AI companion products and production-grade web systems using Next.js, FastAPI, vector memory, Redis, PostgreSQL, and AWS deployment workflows.

  • Took an AI companion product from prototype to production for 500+ active users with memory, messaging, and deployment workflows.

Company projects

NoviAI

Architected and shipped a full-stack AI companion from scratch with conversational AI, memory, and proactive messaging.

Next.js, FastAPI, OpenAI, Pinecone, Llama3, AWS EC2, Redis, PostgreSQL

Projects

2026

PinocchioIndependentGitHub

AI agent framework, CLI engineering

A developer tool for running AI agents with clearer control over planning, editing, tool use, and human review. The focus is making agent work feel predictable: you can see what the agent is doing, steer it when needed, and trust the workflow instead of treating it like a black box.

2025

BetterIndexOpen sourceGitHub

Multi LLM chat app

A chat product for comparing and using multiple AI models from one place. It is built around the user workflow rather than a model list: start a conversation, switch models when useful, compare responses, and keep the interface clean enough for daily use.

Taste

Agent UX

Tools should be visible, debuggable, and recoverable. Agent magic is only useful when the runtime is understandable.

Product engineering

I like interfaces that make operational work faster: fewer screens, clearer states, and fewer hidden assumptions.

Code taste

Small modules, explicit contracts, and boring infrastructure beat clever abstractions when a product has to keep shipping.

Hire me

I am strongest in small teams where the work crosses product, backend, and AI systems. Recent work includes DevOps agents, secure credential access, HITL chat workflows, auth systems, billing, real-time dashboards, and AI products with memory.

FrontendReact.js, Next.js, TypeScript, TailwindCSS, Mantine UI

BackendNode.js, Express.js, FastAPI, Python, WebSockets, BetterAuth

DataPostgreSQL, MongoDB, Redis, Pinecone, Prisma, Drizzle ORM

AI Agents & infraOpenAI, LangChain, Llama3, AWS EC2, Cloudflare, Stripe

singewartanmay@gmail.comGitHubLinkedInResume