Position: Computer Vision / ML Engineer (Intrusion Detection)

Company: NDA ( Gov Project)

Format: contract project / possible long-term cooperation

Location: remote

Salary: discussed individually (depending on experience)

Responsibilities

• Development of an Intrusion Detection System from scratch • Building a complete model training pipeline • Implementation of detection zone logic (polygon-based restricted zones) • Implementation of virtual line crossing logic • Processing object movement directions • Handling cases of partial human visibility • Exclusion of animals and environmental movement • Generation of structured JSON system events • Optimization of models for real-time video processing

Requirements

• Strong Python • Experience with PyTorch or TensorFlow • Experience in developing computer vision models • Experience with detection + tracking systems • Experience with video stream processing (RTSP) • Understanding of inference latency optimization • Experience in developing production-ready code

Will be a plus

• Experience with Torch/ YOLO/ ArcFace / DETR or similar detection models • Experience in optimizing real-time inference • Experience with containerization (Docker) • Experience in developing video analytics systems

We offer

• Participation in the development of an AI video analytics system for government projects • Opportunity for long-term cooperation and expansion of tasks • Expansion of the financial scope upon successful project completion • Opportunity to participate in the development of new platform models • Potential relocation opportunities upon project expansion

Answers to 10 important questions

1. Data: Video streams from CCTV systems (RTSP) + proprietary datasets.
2. Hardware: GPU servers for model training and inference.
3. Business impact scale: Intrusion Detection is a core module of the AI video analytics platform.
4. Data Science maturity level: The team is developing production computer vision models.
5. ML Engineer role: Development of models, system architecture, and production deployment.
6. Manager background: A team of engineers with experience in developing AI security systems.
7. Meeting frequency: Minimal number of meetings, focus on development.
8. Career growth: Upon successful implementation – participation in the development of subsequent models.
9. Prod / Research: Prod ~80%, Research ~20%.
10. Service or leader function: Engineering role with the ability to influence system architecture.

To apply telegram: asinistrapelevin

Please indicate: • relevant computer vision projects • proposed architectural approach • estimated development timeline (3–4 weeks)