Technical Lead - Python, Computer Vision & Edge AI

What you will own

• Delivery and people: Turn requirements into priorities, estimates and acceptance criteria; men-

tor developers across AI, backend, frontend and edge software, review their work and resolve

blockers.

• Architecture: Improve Python services and video pipelines through reusable components, clear

interfaces and configurable site deployments.

• Release quality: Own code review, automated and video-based checks, software/model version-

ing, deployment verification, documentation and rollback.

• AI evaluation: Work with R&D to assess detection, tracking, classification and OCR across

day/night traffic, glare and different camera views; make reports and performance claims

reproducible.

• Reliability and security: Improve camera recovery, offline operation, data synchronisation, mon-

itoring and incident diagnosis; maintain access controls, secrets, dependencies and backups.

• Integration and acceptance: Coordinate camera, RFID, radar and third-party interfaces with

engineers/vendors; support field trials, customer acceptance and technical escalations.

What you bring

• Strong Python, production APIs/backend development, SQL and Linux debugging.
• Experience delivering and supporting a production computer-vision, video-analytics or comparable

edge/IoT system; practical understanding of CV accuracy and runtime performance.

• Git, code review, automated testing, Docker and CI/CD or another repeatable release process.
• Demonstrated mentoring, technical judgement and clear communication of scope, risks and trade-

offs to technical and non-technical colleagues.

• A relevant engineering/computing degree or equivalent practical experience. Strong applicants

outside the stated experience range may be considered.

Useful experience

FastAPI, PostgreSQL, OpenCV, YOLO, ANPR/OCR; NVIDIA Jetson or GPU servers, DeepStream,

GStreamer, TensorRT or CUDA; RTSP/multi-camera systems, networking and remote device manage-

ment; cloud dashboards/frontend integration; ATCC, ATMS, MLFF tolling or industrial automation.

You do not need every listed technology. We value ownership of working systems and the ability to

learn our stack.

First 90 days: Establish a prioritised backlog with owners, strengthen testing and deployment prac-

tices, baseline accuracy/reliability, and take one priority release through testing and verified deploy-

ment.

Location and commitment: Work from our Madurai office; relocation is required for applicants

based elsewhere. We welcome candidates seeking sustained product ownership and a long-term

career with COS AI.Role & responsibilities

Preferred candidate profile

Perks and benefits