Pasha K.

AI/ML Developer | LLMs | Classical ML | Agents

Are you looking for an AI engineer who can take models from idea to production? You're on the right page. Hi, my name is Pasha, I'm an Full Stack Developer with 6.5+ years of experience in AI/ML, specializing in LLMs, AI agents, speech systems, and production-grade AI backends. I design and build intelligent systems that actually work in real products. I've also worked as a Team Lead for an AI engineering team in an agency, and I currently act as a mentor, helping engineers grow and ship high-quality AI solutions. My core competencies include working with: - Python - C++ - C# - AWS/Azure - Groq - FastAPI - TensorFlow - LLMs & Agents Moreover, I am comfortable working both independently and in cross-functional teams, and I focus heavily on clean architecture, reliability, and long-term maintainability of AI systems. More about my skills that will help us achieve the best result: 1. AI/ML Stack: - Large Language Models (LLMs), AI Agents, Tool-using agents - RAG systems (LangChain, LangGraph, LlamaIndex, PGVector) - Voice Agents: Speech-to-Text / Text-to-Speech (Whisper, real-time ASR pipelines, Elevenlabs, OpenAI Realtime API) - Prompt engineering, evaluation, hallucination reduction - Model orchestration, inference optimization, streaming AI - TensorFlow/Pytorch, Python-based ML pipelines, HuggingFace 2. Backend & infrastructure (AI in production) - Python, FastAPI, Django - OpenAI (GPT, Whisper), Groq - PostgreSQL, MySQL - NoSQL: MongoDB, Redis - Vector databases: Pinecone, PGVector, Qdrant - AWS & Azure (deploying scalable AI services) 3. Supporting full-stack experience - React / TypeScript (AI dashboards, internal tools) - REST & streaming APIs for AI systems - CI/CD, Git (GitHub, GitLab, Bitbucket) 4. Leadership & collaboration - Led an AI engineering team in an agency environment - Mentoring developers in AI/LLM systems and production practices - Designing architectures, reviewing code, and setting technical direction - Strong focus on delivery, quality, and long-term maintainability AND: - I enjoy tackling complex AI problems - I learn fast and stay up-to-date with modern AI tooling - I aim to build systems that scale, not prototypes that break If this sounds like a fit, I'd be happy to discuss your project — feel free to invite me to an interview.