Zerodha Claude MCP
MCP server bridging Claude AI with Zerodha Kite — automated trade execution and portfolio analysis.
AI Systems Engineer
I architect Agentic AI systems, developer platforms, and AI-native products at Walmart.

About
AI Systems Engineer with 6+ years building production software and AI platforms at Walmart and Sears. I specialize in architecting Agentic AI systems that combine LLMs, APIs, distributed services, vector databases, orchestration frameworks and modern web applications into intelligent business products. Passionate about designing AI systems that automate real-world enterprise workflows rather than standalone AI demos.
0
Years Exp
0+
Projects
0+
Open Source
Focus Areas
From agentic AI systems and developer platforms to AI-first products — here's where I spend my engineering energy.
Currently Building
Agentic finance assistant for portfolio insights and trade workflows.
Composable AI dashboards that generate and adapt to business KPIs.
LLM-driven Playwright test generation and CI/CD quality gates.
Rapid prototypes exploring AI-native vertical SaaS opportunities.
Highlights
Standout work across AI systems, developer platforms, and AI-native products.
MCP server bridging Claude AI with Zerodha Kite — automated trade execution and portfolio analysis.
Multi-agent orchestration with LangGraph — coordinated LLM agents for complex enterprise workflows.
Agentic RAG evaluation — measure and improve retrieval-augmented generation pipelines.
LLM-powered test generation from user stories — outputs Playwright suites with CI integration.
AI copilot for personal finance — portfolio analysis, risk signals, and conversational insights.
Writing
Deep dives, engineering notes, and blog posts on AI systems, web platforms, and product building.
Headless blog on Next.js and Sanity — technical writing on web development, AI systems, and engineering craft.
Lessons from evaluating retrieval quality, chunking strategies, and agent orchestration in enterprise settings.
Bridging Claude AI with trading APIs and enterprise tools through the Model Context Protocol.
Community
Experiments, demos, and tools shared openly with the developer community.
End-to-end CNN for self-driving using Nvidia methodology — camera preprocessing and Udacity simulator.
Real-time peer video chat with signaling and sockets.
Full-stack ordering flow with real-time status updates.
Career
Software Engineer
Bengaluru, India
Full Stack Developer
Bangalore (Remote)
Technology Consultant
Mohali, Punjab
B.Tech — Computer Science
Have a project in mind or want to collaborate? Send a message — I typically reply within 48 hours.