Skip to content
View 1RaghavM's full-sized avatar
4 followers · 8 following

Achievements

Achievement: Pull Sharkx3Achievement: YOLOAchievement: Pair Extraordinairex2Achievement: Quickdraw

Achievements

Achievement: Pull Sharkx3Achievement: YOLOAchievement: Pair Extraordinairex2Achievement: Quickdraw

Highlights

  • Pro

Organizations

@Opentrons

Block or report 1RaghavM

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
1RaghavM/README.md

Raghav Mehta

I build AI applications and developer tools end to end: Model Context Protocol (MCP) servers, LLM-powered apps, agent backends, and computer-vision models I train and quantize myself.

Open to 2027 AI/ML and software engineering internships, especially in AI applications, developer tools, and backend.

GitHub activity

2,200+ contributions in the past year, including 100+ pull requests.

Raghav's GitHub stats Top languages

Open-source contributions

26 merged PRs to Opentrons/opentrons (lab-automation robotics monorepo) since Sep 2026. Most add runtime-parameter (RTP) support to step-generation command state updates. Recent ones:

Featured projects

Project What it is Stack
visualgen-mcp MCP server that lets Claude Code generate images (Imagen 4) and video (Veo 3.1) straight into a project. Has 6 tools including async video jobs, and CI runs ruff, mypy, and pytest. Python
cpproad 345-lesson interactive C++ course with Judge0-sandboxed grading, a bring-your-own-key AI tutor with 4-tier hints, and spaced review. TypeScript, Next.js, Supabase
dormdish-backend Backend for a college dining app, including a streaming conversational agent with recoverable runs, confirm-before-save, and feature-flagged rollout. Python, FastAPI
FireDetection YOLO11n flame and smoke detector (mAP50 0.945). Its INT8 OpenVINO export is 3.2 MB and runs at 13.6 ms per image. Python, YOLO11
ai-engineering-illustrated-diagrams Illustrated explainers on sandboxes, MCP, inference serving, and agent harnesses, written for my Substack. Diagrams

Also building: ebay-agent (WhatsApp to eBay listing via an MCP server) · arbiter (cost/quality model routing) · network-rag-api (RAG with FastAPI and Ollama) · vision-tune (X-ray VLM fine-tune) · l1-cache (C++ latency experiment)

Stack: Python · TypeScript · C++ · PyTorch · FastAPI · Next.js · Supabase · PostgreSQL · Docker

Get in touch

If you're hiring for 2027 AI/ML or SWE internships, I'd like to hear from you. You can find my writing on Substack and my code here on GitHub.

Pinned Loading

  1. visualgen-mcp visualgen-mcp Public

    Model Context Protocol (MCP) server that lets Claude Code generate images (Imagen 4) and videos (Veo 3.1) via the Gemini API, straight into your project.

    Python 2 2

  2. cpp-course cpp-course Public

    cpproad: interactive C++ course (345 learncpp.com lessons) with sandboxed Judge0 grading, spaced review, and a bring-your-own-key AI tutor.

    TypeScript

  3. dormdish-backend dormdish-backend Public

    FastAPI backend for DormDish, a college dining and nutrition app: campus menus, meal plans, Gemini food scanning, and a feature-gated conversational agent.

    Python

  4. FireDetection FireDetection Public

    YOLO11n fire and smoke detector with an INT8 OpenVINO export (3.2 MB, mAP50-95 0.637).

    Jupyter Notebook

  5. ebay-agent ebay-agent Public

    WhatsApp-to-eBay listing agent: send photos and a sentence, get a sandbox listing, via a custom Model Context Protocol (MCP) server on Hermes Agent.

    Python

  6. ai-engineering-illustrated-diagrams ai-engineering-illustrated-diagrams Public

    Illustrated diagrams explaining how AI systems work: code sandboxes, MCP, LLM inference serving, computer-use agents and agent harnesses. From ragmehta.substack.com.

    Python