AI Integration Engineer

Strong Python: Proven commercial experience writing clean, scalable, and maintainable Python code.

Video Fundamentals: A solid understanding of how video streams work under the hood (codecs, containers, timestamps, keyframes).

Pipeline Integration: Confidence working across complex APIs, third-party libraries, and scalable data pipelines.

Desirables:

Direct experience with FFmpeg or GStreamer.

Familiarity with GPU-accelerated video/AI tech (CUDA, PyTorch, TensorRT, DeepStream).

Experience with multimodal AI (VLMs, LLM APIs, Retrieval-Augmented Generation).

Cloud and containerization skills (AWS, Docker).

Working with a pioneer in advanced video compression, you will build and integrate multimodal workflows that connect cutting-edge video technologies to modern AI ecosystems.

This client would be requiring candidates to do three days on-site per week and two days working remotely.

Core Responsibilities:

Develop production-quality Python pipelines for video processing (ingest, frame extraction, transcoding, and A/V sync).

Integrate proprietary codec SDKs into AI and media frameworks (e.g., FFmpeg, GStreamer, NVIDIA NeMo).

Connect video processing stages to computer-vision and multimodal AI models (for tasks like classification, embeddings, and structured output generation).

Prototype and optimize codec-aware inference workflows to improve accuracy, latency, and cost.