At Geniusee, I lead end-to-end development of Generative AI, Agentic AI, and 3D Computer Vision proofs of concept, from research and model evaluation through AWS deployment. I also contribute to a large production system by developing asynchronous, event-driven backend services.
For Geniusee's Camera Position Generation project, I developed a 3D Gaussian Splatting pipeline for scene reconstruction and novel-view rendering, with image refinement using Qwen-Image-Edit and LoRA. I also built an alternative pipeline combining Gemini camera-aware scene descriptions with Nano Banana for novel-view generation.
At It-Jim, I developed Audio AI and NLP solutions spanning speech processing, audio enhancement, music generation, and LLM-powered agents. For commercial music generation, I led backend development and designed a pipeline that used video content to select and generate music.
In my Audio Intelligence Platform pet project, I built an event-driven audio processing platform with a FastAPI backend and an automated speech pipeline for transcription and speaker diarization. I also developed a LangGraph agent with hybrid RAG and timestamp-grounded responses.

