Role Overview:
- Explored prompt strategies, fine-tuning, and architecture adaptation for clinical note generation.
- Built agentic workflows integrating SFT and instruction-tuned models for steerable outputs in medical contexts.
- Fine-tuned and evaluated LLMs (Phi-3-Vision, GPT-4, Mistral, LLaMA) across enterprise and multimodal retrieval tasks.
- Collaborated on experimental pipelines for multimodal models (Phi-3-Vision, GPT-4o) in Healthcare and Autonomous Driving.
- Developed scalable benchmarking setups for LLM search and retrieval performance.
- Ensured compliance with EU AI Act and MDR regulations to enable trustworthy AI adoption in healthcare.