About Our Client The organisation operates in the software industry, specialising in analytics within the technology and telecoms sector.

- Lead AI Research Scientist (Computer Vision) to architect and scale our next-generation multimodal systems. This is not a standard model-wrapper role; we are looking for a first-principles researcher who can design, pre-train, and fine-tune custom Vision Language Models (VLMs) and advanced computer vision architectures to solve complex, unstructured data extraction and spatial reasoning challenges at scale.

The Successful Applicant

We're looking for someone who has:

- Architectural Ownership: Design, train, and deploy bespoke Computer Vision models, VLMs, and multimodal architectures (e.g., Vision Transformers, Diffusion models, Cross-attention mechanisms) tailored to highly specific domain constraints.
- Applied SOTA Research:



Lead the translation of cutting-edge academic research into production-ready IP. Drive initiatives in zero-shot/few-shot learning, open-vocabulary object detection, and document/scene understanding.
- Performance Engineering & Optimization: Optimize model inference latency and throughput. Implement advanced techniques such as quantization (INT8/FP8), knowledge distillation, model pruning, and FlashAttention to ensure models run efficiently within our cloud infrastructure (AWS) and FastAPI backends. What's on Offer - A permanent role in the technology and telecoms industry - Opportunities to work on creative software solutions - Chances to grow your skills in analytics and computer vision - A collaborative and supportive work environment If this role interests you, please consider applying.

📌 Computer Vision Engineer || WFO || Bangalore (Bengaluru)
🏢 Michael Page
📍 Bengaluru