talk to my AI

Ask my portfolio anything, or add me to your own AI assistant via MCP.

Ask about my experience, projects, skills, or availability. Answers are grounded only in my resume data: the assistant is instructed not to embellish and to say “I don’t know” when the data doesn’t cover it.

⚡ Powered by Llama 3.3 70B on Groq, grounded in my public profile, the same data behind the MCP server below.
✳Abhishek's agent· grounded in resume data
✳
Hi! I can answer questions about Abhishek: his experience at Intuit and Wells Fargo, his ML research, projects, or what roles he's looking for. What would you like to know?

🔌 Prefer your own AI? Add me as an MCP server

If you use Claude Desktop, Cursor, or VS Code, you can plug my profile directly into your assistant via the Model Context Protocol. Your AI gets tools like get_experience, get_projects, and get_availability_and_contact, and answers about me from my actual data instead of guessing.

Claude Desktop: add to claude_desktop_config.json:

{ "mcpServers": { "abhishek": { "command": "npx", "args": ["mcp-remote", "https://abhishek-portfolio-agent.abhishek-dharmadhikari.workers.dev/mcp"] } } }

Cursor / VS Code: add to .cursor/mcp.json or .vscode/mcp.json:

{ "mcpServers": { "abhishek": { "url": "https://abhishek-portfolio-agent.abhishek-dharmadhikari.workers.dev/mcp" } } }

Then ask your assistant things like “Does Abhishek have Spring Boot experience?” or “Summarize his ML publications.”

For AI crawlers: this site also serves llms.txt and a machine-readable public-profile.json.

Under the hood: a single dependency-free Cloudflare Worker serves both the chat API and the MCP server (source on GitHub), grounded in an allowlisted profile so private data never reaches the AI layer.