About Us: We are an engineering team specializing in applied machine vision and video analytics for B2B (industry, logistics, retail). We implement solutions that work in real-world conditions on actual sites: from PPE control on construction sites to transport monitoring at gas stations. Our focus is Edge analytics, operation within closed local networks, and inference optimization.

Job Overview: We are looking for an AI/ML Engineer with a focus on computer vision, who will take ownership of model development from the R&D stage to deployment on customer hardware. You will be working with real-time video streams, solving non-standard problems in object tracking and detection (people, heavy equipment, transport), and extracting maximum performance from local computing resources.

Organizational Structure: In this position, you will work closely with backend developers, the tech lead, and directly with the founders. You will be able to influence architectural decisions of projects and see how your models bring measurable business value.

What a Successful Specialist Looks Like: Success in this role means the ability to think not only about abstract model quality metrics (accuracy/mAP) but also about business outcomes. The ideal candidate is an autonomous specialist who understands that stable 24/7 system operation without internet channel overload and false positives is important for the client. You know how to balance complex modern architectures with inference speed on limited hardware.

Responsibilities:

• Development, training, and fine-tuning of computer vision models (object detection, classification, segmentation, tracking) for real-time video stream processing.
• Optimization of trained models for fast inference on local servers (Edge AI), considering hardware limitations.
• Collection, annotation, cleaning, and augmentation of datasets for specific business scenarios (e.g., construction sites, warehouse terminals, gas stations).
• Integration of ML models into the overall product architecture and development of APIs for interaction with the backend.
• Analysis and elimination of performance bottlenecks in video processing pipelines.
• Participation in technological research (R&D) to select optimal open-source libraries, considering licensing requirements (for implementation in the public sector/Enterprise).

Qualifications:

• Strong Python skills and experience with core ML frameworks (PyTorch - mandatory).
• Deep understanding of modern CV architectures (YOLO v8/v11 family, ResNet, Vision Transformers, etc.).
• Practical experience in model inference optimization (ONNX, TensorRT, OpenVINO).
• Experience with computer vision and video processing libraries (OpenCV, FFmpeg, GStreamer).
• Skills in containerizing ML services (Docker, Docker Compose) and using version control systems (Git).
• Ability to clearly and structurally formulate technical solutions in both code and documentation.

Will Be a Significant Plus:

• Experience in developing and optimizing LLM / RAG architectures (we are also implementing AI assistants).
• Knowledge of C++ sufficient for integrating and accelerating ML models.
• Experience with trackers (DeepSORT, ByteTrack, and similar).

What We Offer:

• No bureaucracy or micromanagement: we value results and an engineering approach.
• Real industrial tasks where your code operates in factories and terminals.
• Remote work, salary discussed based on interview results, flexible schedule