Andrii Siloch

Designed and deployed scalable AI microservices and LLM pipelines for Retrieval-Augmented Generation (RAG), real-time image processing, and meeting summarization. Led the development of a LangChain-based RAG system leveraging FAISS and Pinecone, and engineered scalable FastAPI-based microservices deployed across AWS Bedrock environments with Dockerized architecture.

Built scalable AI backend systems and multimodal pipelines that accelerated personalized video generation and streamlined creative workflows. Architected LLM-driven systems to extract plot elements and emotional nuances from multimedia scripts, and developed a real-time low-latency backend using Django and Channels for collaborative video editing.

Applied machine learning techniques to biomedical data, focusing on gene-based cancer detection and efficient pose estimation tools for medical imaging systems. Contributed to cancer detection research by building early-stage machine learning pipelines from gene expression datasets and assisted in the development of pose estimation models using OpenPose.

Studied computer science at Odessa Polytechnic National University from 2010 to 2014. The curriculum covered fundamental concepts and advanced topics in the field.