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

Block or report pgeedh

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
pgeedh/README.md
Pruthvi Geedh. Robotics engineer building the deployment layer for physical AI.

Website LinkedIn X Substack

Neuracore RCA Robotics Lab Bristol Robotics Lab Univ. of Bristol Univ. of Mumbai

About

Physical AI is advancing fast, but almost none of it reaches a real factory floor. The gap between a model that works in simulation and a robot that works in production is where I work: VLA models, world models, SLAM and navigation. I lead solutions engineering at Neuracore and founded ARTHÉ, an open-source robotics community.

No hype: real deployments, real failures, and what I learned fixing them.

Join ARTHÉ on LinkedIn
Can you operate robots? 3 missions, 3 tasks: a nuclear plant leak with a humanoid, a biohazard lab with two Franka arms, a Mars sample return with a Husky rover. Click to play.

Highlights

Research tooling · LaTeX · AI

The research operating system for robotics. Explore papers, track what you read and write your own in one place, with Notion-style inline AI commands in a LaTeX editor.
Hugging Face · PyTorch · robot learning

Open pull requests to Hugging Face's LeRobot evaluation pipeline: persisting recorded datasets across batched evals and adding provenance metadata to eval results. Also contributing fixes to VLA-Adapter.
Gemini Robotics ER 2.0 · VLA 2.0 · ROS 2

A developer cookbook for Google DeepMind's Gemini Robotics: 31 prompt cards for spatial grounding, grasping, motion planning and safety, a ROS 2 bridge, and the official benchmarks. Translated into 4 languages.
YOLO11 · Gemini · computer vision

A bike-fit studio on your laptop. YOLO11 tracks your pose while you pedal, and Gemini writes an expert-level fit report from the biomechanics.
More projects
Project What it does
GemmaSense Hybrid edge-cloud vision-language system: on-device PaliGemma 2 escalating to Gemini when confidence drops
Open-ENPIRE-Gripper Open design of NVIDIA's ENPIRE compliant gripper, adapted to other robot arms (Robotiq Hand-E, OpenArm)
WorldLab Robotics Examples Photorealistic digital-twin environments for robot training, generated with World Labs
Reachy-Mini · Gemini + PersonaPlex An empathetic desktop robot with Gemini vision, voice wake-words and 15+ physical expressions
mujoco-playground Play three robots in the browser: G1 humanoid, bimanual Franka stacking and a Husky on Mars, scored like a benchmark
UPenn Robotics Specialization Aerial robotics, perception, estimation, planning and mobility coursework

Work with me

Research collaboration, talks and partnerships, or a mentoring call.

Book a call Email me

Pinned Loading

  1. GeminiRobotics-InsightHub GeminiRobotics-InsightHub Public

    Python 3

  2. WorldLab-RoboticsExamples WorldLab-RoboticsExamples Public

    WorldLab Robotics Examples is the definitive resource for bridging the gap between Generative AI and Embodied Intelligence.

    Python 5

  3. Open-ENPIRE-Gripper Open-ENPIRE-Gripper Public

    Democratizing the NVIDIA ENPIRE compliant gripper design for diverse robot arms and physical AI research.

    Python 5

  4. mujoco-playground mujoco-playground Public

    Real Unitree G1 humanoid in real MuJoCo physics, jumping over obstacles

    JavaScript