Skip to content
View MadhankumarAI's full-sized avatar
🎯
Focusing
🎯
Focusing

Block or report MadhankumarAI

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.

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
MadhankumarAI/README.md

Hey there 👋, I'm Madhan Kumar R

AI & ML @ DSCE · GSoC 2026 · Generative AI Intern @ AllCognix AI · Co-Founder & CTO @ Lumus

Typing SVG


About Me

  • 🎓 B.E. in AI & ML @ DSCE — CGPA 9.89
  • 🛰️ GSoC 2026 Contributor @ InVesalius — TotalSegmentator for CT/MRI segmentation
  • 🧠 Generative AI Intern @ AllCognix AI — building distributed data pipelines for agentic AI
  • Co-Founder & CTO @ Lumus — AI for non-surgical facial enhancement
  • 🔒 Patent holder — AeroDrive, wind-powered EV charger
  • 👨‍🏫 Mentored 250+ students in MERN @ JIT Davangere
  • 🤝 Open to collaborations on agentic AI, medical imaging, and edge AI research
  • ⚡ Fun fact: I spend equal time on PyTorch and GLSL shaders

🚀 What I'm Building Right Now

🛰️ TotalSegmentator → InVesalius (GSoC 2026)

Integrating a lightweight TotalSegmentator pipeline into InVesalius for automated anatomical segmentation of CT & MRI volumes. Building the inference pipeline with PyTorch, NumPy, and wxPython — optimized for low-memory clinical workstations while preserving multi-organ accuracy.

🧠 Data Studio @ AllCognix AI

Building a platform that turns raw multi-modal data into AI-ready datasets — ingestion, cleaning, annotation, and export pipelines. Architecting the stack with Apache Spark, Kafka, Airflow, Delta Lake, DVC, and Label Studio. Data versioning, lineage tracking, and human-in-the-loop annotation throughout.

⚡ Lumus (Co-Founder & CTO)

Co-founded Lumus, an AI platform for non-surgical facial enhancement trained on insights from 2000+ peer-reviewed medical papers. Owning the end-to-end CV pipeline — MediaPipe, Vision Transformers, fine-tuned BiSeNet, FastAPI backend — and architecture decisions through MVP.


🛠️ Skills & Tools

Languages

AI / ML

Also working with HuggingFace, LangChain, MediaPipe, Vision Transformers

Web / Backend

Plus React Three Fiber, GLSL shaders, Framer Motion

Data / Infra

Plus Apache Spark, Airflow, Delta Lake, DVC, Label Studio

Tools & Platforms


📈 GitHub Stats


🏆 Selected Achievements

  • 🥇 1st Place — AIverse (Amrita) 2026 · Docothon (CellStrat) 2025
  • 🥈 1st Runner-Up — Dev Dual, IIT Hyderabad 2026
  • 🔒 Patent PublishedAeroDrive (App. No. 202441090880 A)
  • 🌟 GSoC 2026 Contributor @ InVesalius
  • 🧑‍🏫 Lead, Code Zero Club @ DSCE
  • 💼 Python Development Intern @ Zaalima Dev Pvt. Ltd. (Mar–June 2025)
  • 🛠️ Upstream contributions accepted into FastAI

🌱 Currently Exploring

  • Agentic AI workflows + distributed training with Ray
  • Medical imaging research (segmentation, multi-modal CT/MRI)
  • Novel LLM architectures — Mixture of Recursion, Contextual Sparsity
  • Real-time 3D for web — Three.js, React Three Fiber, GLSL

💬 Let's Connect

Portfolio LinkedIn GitHub


"Greatness is never rushed. It takes time, patience, and relentless hard work."

Pinned Loading

  1. Dockothon Dockothon Public

    JavaScript 3

  2. InterXAI InterXAI Public

    Forked from RITVIKKAMASETTY/Interxaii

    JavaScript

  3. MORS MORS Public

    Python

  4. Storynarrator_app Storynarrator_app Public

    Dart

  5. invesalius/invesalius3 invesalius/invesalius3 Public

    3D medical imaging reconstruction software

    Python 795 435