Gen AI Solution Architect
augļus
Feb 2024 - Oct 2025 (1 year 9 months)
- Designed and implemented multimodal AI systems integrating vision and language models for advanced skincare diagnostics.
- Developed robust retrieval-augmented generation (RAG) pipelines combining LLMs (e.g., Llama‑2 13B) and vector search (using FAISS) to improve diagnostic accuracy.
- Solved complex GPU memory challenges through advanced offloading and quantization strategies, ensuring efficient deployment of large-scale models.
- Engineered scalable solutions for real-time inference in dermatology, leveraging cutting-edge deep learning and computer vision techniques.
- Created user-friendly interfaces with Gradio to streamline end-to-end AI experiences and facilitate interaction with AI-driven diagnostic tools.