AI Systems Engineer & Developer Advocate with ~5 years building production AI agents, verifiable ML pipelines, and developer ecosystems at the intersection of Gen AI · Web3 · Edge ML.
AI Agent Hub & marketplace enabling real-world workflow automation. Combines crypto agents (blockchain transactions, NFT minting via natural language), browser automation, and a multi-LLM orchestration backend. USDC escrow and reputation system built-in for developer monetization.
UPI middleware for zero-fee collections that verifies customer payments made directly to merchant bank accounts, then orchestrates trusted post-payment fulfillment events without acting as a money-moving processor.
AI-native conversational commerce platform using RAG + multimodal models to search, compare, and purchase across multiple e-commerce sites using natural language.
Integration of LLaMA into RISC-V architecture enabling deterministic, verifiable LLM computation on-chain. First-of-kind implementation for auditable AI inference.
Mobile-powered respiratory health monitoring with on-device ML inference. TensorFlow Lite model detecting coughs, wheezing, and breathing anomalies at 200–300ms latency.
Integration tool combining Cartesi's verifiable compute with decentralized identity (DID), enabling on-chain identity verification workflows. Achieved strong adoption among builders.
A kernel-based operating system for AI agents with safety-first design. Provides robust, secure, and scalable platform for running AI agents with centralized control, system call interface, and fine-grained permission system.
Confidential AI via TEEs — privacy-focused LLM infrastructure that uses Trusted Execution Environments to provide cloud-based inference with strong confidentiality guarantees. End-to-end encrypted with X25519 key exchange and AES-GCM.
Decentralized Services Marketplace — a blockchain-powered freelance platform with escrow payments, AI integration, and dispute resolution. Built with Next.js, Node.js, Prisma, PostgreSQL, and Solidity on Base.
Agentic Market Coordination Layer — a decentralized marketplace where autonomous AI agents discover tasks, compete through bidding, execute work, and receive payment via smart contracts. Human-in-the-loop review with trustless escrow payments.
AI-Powered Resume Builder & Mock Interview Platform — combines intelligent resume optimization with role-specific AI mock interviews. Built with Next.js 16, React 19, TypeScript, Tailwind CSS v4, and Anthropic Claude API.
I'm a builder at the intersection of AI systems, developer experience, and emerging compute paradigms. My career has been defined by one thesis: the most impactful work happens where complex technology meets the humans who build with it.
At Cartesi, I turned a 2–3 hour developer pain point into a 15-minute workflow — not through clever code alone, but through systematic feedback loops, empathy, and cross-functional execution. That's the kind of 10x I care about.
Today I'm focused on the hardest problem in AI: making increasingly powerful models observable, debuggable, and trustworthy. I think about interpretability as an engineering discipline — not an academic one — and I design production systems with that constraint built in from day one.
IIT Roorkee physics graduate. Shipped hardware to 35,000+ customers. Spoken at ETHGlobal. Currently building AI agent infrastructure that actually works.
Open to AI engineering, developer advocacy, and technical leadership roles. Especially interested in teams working on LLM evaluation, agentic systems, or AI safety tooling.