I'm a Computer Science & Data Analytics undergrad at IIT Patna, interested in building production-grade systems at the intersection of AI, backend engineering and financial markets.

Previously, I worked in the Founder's Office at Multibagg AI, where I helped build and scale AI-powered stock research systems used by real investors.

• 🤖 Built agentic workflows for Ask Iris, a multi-agent stock research assistant that has handled 500K+ user queries
• 📚 Worked on retrieval and RAG systems spanning 100K+ documents and 20M+ records
• 🧪 Built evaluation infrastructure with LLM-as-a-Judge, tool-call validation, citation checks, latency tracking and failure analysis
• 📈 Engineered financial data, IPO, ETF and market intelligence pipelines
• ⚙️ Worked across TypeScript, Node.js, Python, PostgreSQL, Redis, BullMQ, Qdrant, Pinecone and Docker
• 🧠 Currently exploring AI agents, memory systems, agent harnesses, quant systems and backend infrastructure
• 💼 Looking for AI Engineering, LLM Engineering, FDE and Backend/SWE opportunities

📍 India 📩 [email protected] 🌐 Portfolio 📄 Resume

Agentic AI contributor that scans approved Go repositories, identifies issues, generates patches, validates changes and prepares PR drafts through a guarded multi-stage workflow.

TypeScript Next.js OpenAI GitHub API Zod

Research-grade paper trading arena where multiple LLMs compete on Nifty 50 stocks using deterministic market snapshots, quant features, execution rules and portfolio-level risk constraints.

TypeScript Node.js Redis BullMQ LLMs Quant

AI-powered system for analyzing public Instagram accounts, ranking high-signal content and extracting creative intelligence from captions, metrics, transcripts and video frames.

TypeScript Next.js OpenAI FFmpeg SSE Zod

Experimenting with persistent memory infrastructure for coding and AI agents, with a focus on giving agents reusable context across sessions and tasks.

AI Agents Memory Agent Infrastructure

A finance-focused tool for understanding margin trading facility economics, leverage, costs and position outcomes.

Finance Trading Web

Founder's Office · January 2025 - June 2026

Worked directly with the founder across AI engineering, backend infrastructure, retrieval systems, financial data and product workflows.

Some of the systems I worked on included:

• Ask Iris - Multi-agent stock research assistant serving 500K+ user queries
• Tool-calling across SQL, RAG, web search and financial data
• Retrieval over 100K+ documents and 20M+ records
• Qdrant/Pinecone infrastructure with hybrid search, reranking, filters, chunking and citations
• LLM evaluation and agent harness infrastructure
• IPO and ETF document intelligence pipelines
• AI-powered stock screening
• Production financial analytics covering 300+ ratios and indicators
• Redis-based infrastructure, queues and automation systems
• Real-time AI systems for market news and exchange announcements

• 🥇 National Finalist - India FinTech Forum FinTech Olympiad 2024
• 🤖 Helped build AI systems serving 500K+ user queries
• 📚 Worked with retrieval infrastructure spanning 100K+ documents
• 🗃️ Built systems operating across 20M+ financial records
• 📊 Maintained financial analytics infrastructure covering 300+ ratios and technical indicators
• 🚀 Built projects across AI agents, quant, financial infrastructure and developer tooling

I'm always up for conversations around AI agents, LLM systems, backend engineering, quant, financial markets and startups.