I build intelligent, end-to-end AI systems, focusing on translating complex concepts into practical, scalable products. My expertise lies in creating robust agentic systems, sophisticated RAG pipelines, and multimodal AI applications.
Recently, I architected and led the development of a multi-agent virtual assistant for a major enterprise in the healthcare sector. The system integrates intelligent chat, video consultations, and a real-time management dashboard. My contributions spanned the full project lifecycle:
My key contributions included:
-
🧠 System Architecture: Designed a multi-agent framework using LangGraph to handle distinct tasks like triage, video calls, consultation, and information retrieval.
-
📚 Data Intelligence: Engineered a high-performance RAG pipeline with ChromaDB to provide agents with accurate, context-aware information.
-
⚡ Real-Time Backend: Built the WebSocket infrastructure to power the live chat and call center monitoring dashboard.
-
🗂️ Data Modeling: Designed the database schemas for managing user sessions, appointments, and analytics.
-
🐳 Deployment: Implemented a fully Dockerized architecture for the FastAPI backend, Ollama model serving, and Next.js frontend.
-
🎨 Full-Stack Development: Owned the complete build, from the Python backend to the Next.js/Tailwind user interface.
-
Agentic AI Systems: Architecting multi-agent frameworks for complex, automated workflows.
-
Retrieval-Augmented Generation (RAG): Building robust pipelines that ground LLMs in external knowledge.
-
Multimodal AI: Integrating models like PaliGemma and CLIP to process and understand both text and images.
-
End-to-End AI Products: Bridging the gap between a Python-based ML backend and a polished, scalable web application.
-
MLOps & Deployment: Containerization with Docker, CI/CD, and cloud-native deployment.
When I'm not architecting AI systems, you can find me performing as a live-coded musician under the name Aviara. I create algorave sets by combining Strudel (for music) and Hydra (for visuals).
I'm also deeply exploring the intersection of human-computer interaction, with a keen interest in Brain-Computer Interfaces (BCI) and the future of neurotechnology.