Orangecells Innovation Lab Pvt. Ltd.

Computer Vision, AI & Backend Engineer Video Intelligence & Analytics Platform Development About the Role We are building a high-scale, real-time video intelligence and analytics platform that processes live feeds to detect people, vehicles, anomalies, crowding, loitering, restricted-zone breaches, etc., and generate instant alerts, searchable clips and in-depth analytics and reports. You will own the entire pipeline end-to-end: from frame ingestion -> analysis -> inference -> learning -> storage -> API exposure -> deployment & monitoring -> long term maintenance. Responsibilities Design and implement real-time video processing pipelines using agentic AI frameworks, OpenCV + Deep Learning models on live streams (20-100+ cameras per tenant) Build high-performance, async FastAPI services (REST + WebSockets) for configuration, live preview, event retrieval, reporting Optimize performance, inference latency and GPU/CPU utilization (frame skipping, ROI cropping, batching, model quantization) Develop event deduplication, tracking, re-identification, and crowd/loitering/altercation logic Design scalable MySQL schemas (or TimescaleDB/PostgreSQL) with proper indexing & partitioning for millions of events/day Containerize everything with Docker, deploy on cloud (AWS/GCP/Azure) or on-prem servers, and set up CI/CD Implement monitoring, logging, alerting, auto-scaling, and graceful degradation when streams drop Write production-grade, multi-threaded Python code that can run 247 without memory leaks Complete ownership of development, quality assurance, maintenance and uptime for the entire backend Must-Have Technical Skills & Experience Crunch time startup mindset Solid grasp of data structures & custom algorithms design Deep knowledge of agentic AI Strong Python proficiency with 5+ years of real-world project experience Hands-on experience with OpenCV for live video processing (frame extraction, resizing, color conversion, ROI, optical flow, etc.) Practical deep-learning experience in object detection/tracking (YOLO family, DeepSORT, ByteTrack, or custom models) on video streams FastAPI (or strong Flask + async experience) building production APIs & WebSockets Database schema design, indexing, query optimization, handling 1M+ rows/day, data backups, data pruning Docker & docker-compose; comfortable writing multi-stage Dockerfiles for OpenCV Linux server administration (systemd, nginx, log rotation, SSH, tmux, resource monitoring, shell scripting, cron jobs) Edge device deployment and management Hands on experience managing production infrastructure Hands on experience with monitoring, alerting and high availability design for production services Hands on experience of designing systems: To be scalable to handle streams from 100s of cameras, To be highly optimized on compute resources, To handle single tenant / multi tenant design, To optimize and reduce storage costs to the maximum, To be extensible and maintainable long term Good-to-Have (Big Plus) Previous work on video/analytics products, ANPR, crowd analysis, or perimeter security Previous work on production infrastructure management Experience with message queues (Kafka, RabbitMQ, Redis Streams) for decoupling ingestion & processing Education BE/BTech/MTech in Computer Science, AI/ML, or related fields Candidates with strong projects and broad background including infrastructure management will be preferred Why Join Us? Opportunity to play a pivotal role in shaping the companys growth trajectory. Opportunity to develop innovative domain-specific solutions from the ground up. Freedom, flexibility, and recognition for results-driven performance.