Computer Vision Engineer – L1
- Led design and deployment of UVSS modules (car model detection, automatic masking, false positive reduction) improving accuracy and reliability.
- Built depth estimation, stereo calibration, and 3D reconstruction pipelines for precise vehicle profiling under real-world constraints.
- Managed client–server multi-model deployments, CUDA optimization, and real-time performance tuning.
- Mentored interns and juniors on depth estimation, synthetic data generation, point cloud refinement, and debugging (e.g., memory leaks).
- Product strategy contributions across hardware–software trade-offs and integration feasibility for next-gen UVSS.
- Recruitment support: resume shortlisting, interviews for ML and C++ roles, planning induction.
Impact: Reduced false positives by 20–25%, Increased threat detection rate by 15–20%; Managed deployment of 3+ models across multiple clients.