ALEKSANDR N.

I build production-ready AI applications that help businesses automate workflows, work with their internal knowledge, and reduce manual work. My main focus is LLM applications, AI Agents, RAG systems, and custom LLM solutions. I can take an AI idea from an initial concept to a working prototype or production-ready application. What I can build for you AI Agents & Agentic Applications * AI agents that use tools, APIs, databases and external services * Multi-step agentic workflows with LangGraph * MCP integrations and MCP-based tool usage * Business process automation * AI assistants that can actually perform tasks, not just generate text RAG & AI Chatbots * Document-based AI assistants * RAG over PDFs, websites, databases and internal documentation * Semantic and hybrid search * Vector databases and retrieval pipelines * Evaluation and optimization of RAG quality * Private enterprise knowledge assistants LLM Applications * Custom AI assistants and copilots * LLM-powered backend services and APIs * OpenAI and open-source LLM integrations * Structured outputs and tool calling * AI-powered automation and internal tools Local & Private AI * Run LLMs locally using Ollama or vLLM * Open-source model deployment and optimization * Reduce dependency on expensive API calls * Keep sensitive business data inside your infrastructure Datasets & Fine-tuning * Dataset preparation for LLM training * Instruction and conversational datasets * Data cleaning and preprocessing * LoRA / QLoRA fine-tuning * Fine-tuning open-source LLMs with Unsloth Why work with me? I have 10+ years of software development experience and extensive experience in machine learning, NLP and LLM-based applications, including work in a large financial organization. I don't approach AI projects as pure research. My goal is to understand the business problem, choose the simplest effective architecture, and deliver a working solution that can actually be used. I'm comfortable working directly with clients: discussing requirements, designing the architecture, making technical decisions, and implementing the solution end-to-end. Tech Stack Languages: Python LLM / AI: OpenAI, Hugging Face, Ollama, vLLM, Unsloth Agents: LangGraph, LangChain, MCP, LlamaIndex RAG: Qdrant, FAISS, ChromaDB, hybrid search Backend: FastAPI, Docker Cloud: AWS AI Development: rapid prototyping, AI-assisted development, vibe-coding workflows If you have an AI idea, an existing prototype that needs to be improved, or a business workflow you'd like to automate, send me a message and let's discuss how to build it.