FADA is seeking a Senior AI Developer to design and deploy AI and machine learning capabilities for geospatial intelligence across cloud and air-gapped environments. The role includes developing deep learning models, fine-tuning foundation models, building agentic AI systems, and supporting mission operations.
Responsibilities:
- Create models for object detection, segmentation, image translation, point cloud classification, time-series forecasting, and vision transformers.
- Fine-tune geospatial and general foundation models using methods such as LoRA and QLoRA.
- Build LLM and VLM workflows and develop agent systems for reasoning, planning, and tool use.
- Generate geospatial embeddings, build knowledge graphs, and develop NLP pipelines.
- Evaluate architectures and support AutoML initiatives.
- Build AI pipelines on Azure and deploy models using Docker, Kubernetes, Helm, vLLM, Triton, TorchServe, FastAPI, Ollama, and LocalAI.
- Optimize models with ONNX, TensorRT, quantization, and distillation.
- Produce architecture documentation and establish MLOps practices such as versioning, tracking, retraining, and monitoring.
- Adhere to responsible AI practices, ensure explainability, and comply with defense requirements.
Qualifications:
- Degree in a relevant field.
- 5 or more years of Python deep learning experience.
- Strong skills in PyTorch, TensorFlow, scikit-learn, NumPy, and Pandas.
- Experience with foundation models, geospatial models, agentic AI frameworks, Azure deployment, offline model hosting, NLP tasks, and API development.
- Expected to demonstrate strong engineering practices.
Preferred Skills:
- Experience with RAG, vector databases, geospatial embeddings, knowledge graphs, open-weight model deployment, RLHF or DPO, GeoAI, satellite imagery, responsible AI, GPU acceleration, distributed training, model optimization, multisource geospatial data, GIS tools, and defense or national security experience.
Work Conditions:
Full-time, on-site in the United Arab Emirates.