Source description
About the role
: Computer Vision Engineer Location: Gurugram Experience: 27 years (we hire for capability, not seniority) About Us Stupa builds cutting-edge AI for real-time sports intelligence ; automated commentary, player tracking, non-contact biomechanics, ball trajectory, LED graphics, and broadcast-grade stats. Your models will be seen live by millions across global events. Global Travel Work that literally travels the world. Youll deploy systems at international tournaments across Asia, Europe, and the Middle East, working inside world-class stadiums, courts, and TV production rooms. What Youll Build AI Language Products Automated live commentary (LLM + ASR + OCR), real-time subtitles, AI storytelling. Non-Contact Measurement (CV + Tracking + Pose Estimation) Player velocity, footwork, acceleration, shot recognition, 2D/3D reconstruction, real-time edge inference. End-to-End Streaming Pipelines Temporal segmentation, multi-modal fusion, low-latency edge + cloud deployment. What Youll Do Train and optimise ML/CV/NLP models for live sports, build tracking & pose pipelines, create LLM/ASR-based commentary systems, deploy on edge/cloud, ship rapid POCsproduction, manage datasets & accuracy, and collaborate with product, engineering, and broadcast teams. Requirements Core Skills Solid ML fundamentals (NLP/CV/multimodal) PyTorch/TensorFlow, transformers, ASR or pose estimation Data pipelines, optimisation, evaluation Deployment (Docker, ONNX, TensorRT, FastAPI, K8s, edge GPU) Strong Python engineering Bonus: Sports analytics, LLM fine-tuning, low-latency optimisation, prior production ML systems. Why Join Us Your models go LIVE in global sports broadcasts International travel for tournaments High ownership, zero bureaucracy Build Indias most advanced AI Sports product Cool, futuristic problems + freedom to innovate Up to 40LPA for exceptional talent You Belong Here If You. Want to build what the world hasnt seen Want impact on live sports Thrive in fast-paced ownership-driven environments. .