Sergii G.

I build production-oriented AI applications using Python, LangChain, AWS Bedrock, Anthropic Claude, FastAPI, PyTorch, and OpenCV. My experience covers LLM-powered applications, Retrieval-Augmented Generation (RAG), AI agents, voice AI, computer vision, medical image analysis, and multimodal machine learning. With 4+ years of software development experience and a strong Python background, I help businesses turn AI concepts into functional applications, from model development and evaluation to API integration. LLM, RAG & AI Agents RAG pipelines using LangChain, PostgreSQL, pgvector, and vector databases. AWS Bedrock integration with Anthropic Claude and Amazon Titan Embeddings. Enterprise knowledge-base assistants and multi-tenant AI applications. Document processing and conversational question answering. Prompt engineering, contextual retrieval, and AI response guardrails. Computer Vision & Deep Learning Image classification and deep learning with PyTorch and OpenCV. Medical image analysis using Fundus and OCT images. Multimodal models combining image, text, and structured data. Model training, validation, evaluation, and explainability using Grad-CAM. EfficientNet, CNNs, and image preprocessing pipelines. Voice AI & Conversational Applications AI voice agent development using Retell AI and Anthropic Claude. Real-time audio streaming and WebSocket integrations. Twilio integration and bilingual conversational workflows. AI-powered call analytics, transcription processing, and automated workflows. Python Backend & AI Integration Python | FastAPI | Django | PostgreSQL | SQLAlchemy | Docker | REST APIs | WebSockets | pgvector I develop AI application backends, integrate ML models into existing products, and implement asynchronous processing pipelines.