Skip to content
Sangam Gupta

Use my portfolio from your AI assistant

This site runs a Model Context Protocol server. Add it to Claude, ChatGPT, Cursor or any MCP client, and your assistant can look up my projects, search my code, check a job description against my work and fetch my resume, all from the same content as this site.

Server address

https://www.sangamgupta.in/mcp

No sign-in. Read-only: nothing an assistant does here can change the site.

How to connect

  • Claude: open Settings, then Connectors, add a custom connector and paste the address above.
  • ChatGPT: add it as a custom MCP connector (in developer mode, where your plan offers it) with the same address.
  • Cursor, VS Code and other MCP clients: add a remote server with this config.
{
  "mcpServers": {
    "sangamgupta": { "url": "https://www.sangamgupta.in/mcp" }
  }
}

Then ask something like “Using the sangamgupta tools, how does Sangam’s work fit this job description?” and paste the role.

What it can do

get_profile
Who Sangam Gupta is: role, location, availability, contact, links, skills, experience and education.
list_projects
Every project with a one-line outcome, status (Live, Code or In progress), stack and case-study URL. Flagship work first.
get_project
The full case study for one project: problem, approach, result, metrics, links, architecture and, for ML projects, the model card.
search_portfolio
Search Sangam's site and the code in his public GitHub repos. Returns the most relevant passages with links. Use it for specific questions, such as which projects used FastAPI or how a model was evaluated.
check_role_fit
Compare a job description with Sangam's documented work. Returns each key requirement as strong, partial or gap, with evidence and a link. Gaps are reported as gaps.
get_resume
His resume as JSON Resume (jsonresume.org), plus the PDF link.
book_call
How to book a call with Sangam. Returns his booking link and email. It does not book anything itself; share the link with the user.

Try it here

Sends a real request to the server and shows the raw JSON-RPC reply, the same thing an assistant receives.