Senior AI/ML Engineer (Computer Vision, Multimodal, LLM, PyTorch

Senior AI and ML Engineer for Computer Vision, Multimodal, LLM, and PyTorch About the role We are an AI product company building a multimodal perception engine that fuses video, image, voice, and text into one conversational, uncertainty-aware pipeline. We are looking for a senior AI and ML engineer to own the research and development of the core models across computer vision, conversational NLP, and multimodal fusion. This is hands-on research and development. You will design, build, fine-tune, and quantitatively validate the models that power the engine. What you will do - Build the multimodal fusion pipeline that combines video, image, voice, and text. - Develop the computer vision components, including segmentation, recognition, and portion or 3D estimation. - Build the conversational, uncertainty-aware disambiguation layer using LLM and Transformer fine-tuning. - Adapt open multimodal foundation models through transfer learning, rather than training from scratch. - Define and run quantitative validation against measurable accuracy targets. Critical skills, must have - Strong Python and deep learning with PyTorch. - Computer vision for image and video, including segmentation, detection, and recognition. - Multimodal AI and vision-language models that fuse multiple modalities into one model. - NLP and LLM fine-tuning with Transformers and Hugging Face. - Experience adapting and fine-tuning open foundation models through transfer learning. Good to have - Video analysis and temporal models, 3D reconstruction, or monocular depth and volume estimation. - ASR and speech-to-text, for example Whisper. - Uncertainty quantification and model calibration. - Applied statistics and data science, including time-series correlation and multiple-comparison control. - MLOps, including experiment tracking, model serving, and cloud GPU on AWS or GCP. Requirements - Residency in an EU country is required. - A PhD in AI, computer vision, machine learning, or a closely related field is a big plus. - Able to work independently and own the model research and development end-to-end. To apply Briefly describe one multimodal or computer vision project you led, including the problem, the models and frameworks you used, and your specific role. Links to a portfolio, GitHub, or papers are welcome.