About the Project
We are building a retail computer vision platform that integrates IP cameras, live video streaming, AI inference, and real-time analytics.
Our platform processes live camera streams from multiple stores, runs face and visitor analytics through GPU inference pipelines, and delivers scalable event-driven backend services designed for high-throughput, multi-site deployments.
Tech Stack
Python, FastAPI, Go, Kafka, Docker, Docker Compose, PostgreSQL, pgvector, Redis, MinIO, MediaMTX, Triton Inference Server, NVIDIA GPU Infrastructure.
Responsibilities
- Develop, maintain, and optimize backend services built with Python and FastAPI.
- Design and improve backend services for multi-store, high-concurrency, real-time systems.
- Maintain and improve event-driven pipelines using Kafka and Go-based workers.
- Integrate and troubleshoot RTSP/IP cameras, edge agents, and live video streaming sessions.
- Work with RTSP, RTMP, HLS, WebRTC, and MediaMTX streaming infrastructure.
- Maintain services using PostgreSQL, SQLAlchemy, Alembic, Redis, and MinIO.
- Optimize database queries, indexing strategies, caching, and data access patterns under production load.
- Design systems that remain stable under high numbers of concurrent streams, events, and inference requests.
- Integrate AI inference pipelines using Triton Inference Server.
- Debug and resolve production issues across application code, containers, networking, databases, message queues, storage, and GPU inference services.
- Improve system reliability, observability, logging, monitoring, retries, timeouts, and fault tolerance.
- Read existing production code, understand business workflows, and safely deliver fixes without disrupting live systems.
- Strong backend engineering experience with Python.
- Production experience with FastAPI or similar backend frameworks.
- Strong experience with Docker and CI/CD.
- Experience with PostgreSQL, SQLAlchemy, migrations, indexing, query optimization, performance tuning, and database troubleshooting.
- Experience with Kafka or similar event-driven systems.
- Strong Linux debugging and production troubleshooting skills.
- Strong understanding of HTTP APIs, WebSockets, asynchronous programming, concurrency, and distributed systems.
- Strong understanding of scalability, bottleneck analysis, resource optimization, and high-throughput backend design.
- Ability to identify and resolve performance issues under production load.
- Ability to read, debug, and maintain Go services, especially workers and consumers.
- Experience with object storage such as MinIO or S3.
- Ability to diagnose and resolve issues across software, infrastructure, networking, and external services.
Important Extra Skills
- Experience with RTSP cameras, FFmpeg, GStreamer, RTMP, HLS, WebRTC, or MediaMTX.
- Experience integrating and operating Triton Inference Server, GPU-based inference services, and NVIDIA tooling.
- Practical experience integrating computer vision pipelines, face embeddings, similarity search, and image preprocessing workflows.
- Experience with pgvector or vector similarity search.
- Experience with Redis pub/sub or distributed worker coordination.
- Experience with production monitoring, logging, and observability.
- Able to work independently on production systems.
- Comfortable working with large existing codebases written by other engineers.
- Able to troubleshoot complex distributed systems methodically.
- Able to build systems that remain stable under growing numbers of cameras, stores, users, and real-time events.
- Communicates clearly, takes ownership, and delivers production-safe solutions
contact:@Karam_Mahfod on telegram