Balaji Sundareshan

Senior Machine Learning Engineer at TuMeke Ergonomics

Computer VisionMachine Learning3D Scene ReconstructionModel QuantizationReal-time TrackingEdge AI Deployment

About

I'm Balaji Sundareshan, a Senior Machine Learning Engineer at TuMeke Ergonomics in the San Francisco Bay Area. My career has been dedicated to the intersection of Computer Vision, Robotics, and AI, moving from research at IIT Delhi and Northeastern to leading technical developments in the startup world. I am deeply passionate about mission-driven work, specifically leveraging AI to improve workplace safety and ergonomics—a focus that has already impacted over 30,000 users in the aviation industry. Beyond my technical work in 3D scene reconstruction and model optimization, I'm an active participant in the Silicon Valley startup ecosystem, frequently serving as a judge for pitch competitions. I'm always happy to share my expertise in research-to-product pipelines and model deployment on edge hardware with fellow innovators and founders.

Networking

What I can offer

›Deep technical expertise in CV/ML architecture
›Model optimization for production (TensorRT, CoreML)
›Research-to-product pipeline strategy
›Startup pitch judging and technical evaluation

Looking for

›expanding my professional network
›exploring mutual opportunities in the San Francisco/Silicon Valley tech ecosystem

Best fit for

Early-stage AI foundersStartup investorsComputer Vision researchersWorkplace safety professionals

Current Interests

Egocentric visionWorkplace ergonomicsStartup ecosystem and pitch competitionsTransformer-based action segmentationReal-time system optimization

Background

Career

Progressed from Computer Vision Researcher at IIT Delhi to Lead Research Engineer at edgetensor, followed by a Research Assistantship at Northeastern and a role at Motional, before joining TuMeke Ergonomics.

Education

M.S. in Electrical and Computer Engineering from Northeastern University; B.Tech. in Engineering Physics from IIT Delhi.

Achievements

›Recipient of the O-1A visa for Extraordinary Ability
›Co-inventor on U.S. patent for 3D object detection fusion
›Developed ergonomic app used by 30,000+ flight attendants
›Improved 3D object detection velocity prediction by 40% at Motional
›Achieved 20x speed-up in model inference using TensorRT
›Top 13 finalist in Mahindra Rise Driverless Challenge

Opinions

Innovation is driven through candid feedback and mentorship for early-stage founders
Strong alignment with mission-driven work, specifically using AI to eliminate workplace injuries