AI Engineer | LLMs | Agents | RAG | Infrastructure Optimization
Building scalable, cost-efficient, and secure AI systems.
I’m an AI Engineer passionate about building intelligent, scalable, and secure AI systems. With a full-stack background, I integrate advanced AI into robust web and mobile applications, ensuring high performance, seamless UX, and secure deployment on private infrastructure. My focus is on developing cost-effective and production-ready AI workflows for startups and enterprises alike. I quickly learn new concepts and keep up to date with the latest advancements in the AI field.
Key Areas of Expertise:
- Large Language Models (LLMs): GPT (OpenAI), LLaMA (Meta), Claude (Anthropic), open-source (Hugging Face) and others.
- Retrieval-Augmented Generation (RAG): Designing and integrating RAG pipelines to enhance context-aware responses.
- AI Agents & Multi-Agent Systems: Building intelligent agents capable of tool use and complex reasoning.
- Fine-Tuning & Prompt Engineering: Customizing models for specific tasks and optimizing prompt strategies.
- Model Optimization & Cost Efficiency: Improving performance and reducing infrastructure costs on private and cloud environments.
- Python 🐍: Core language for AI development, scripting, and prototyping.
- TensorFlow & PyTorch: Training, fine-tuning, and deploying deep learning models (NLP, vision, agents).
- FastAPI: Building high-performance APIs for serving ML models and AI agents.
- SQL: Managing and querying structured datasets for training and evaluation.
- LangChain: Orchestrating LLM pipelines, agents, and RAG systems.
- Custom Tools: Building vector search, embedding pipelines, and inference wrappers.
- Docker
- AWS Cloud Computing ☁️
- Frontend: React ⚛️ | Astro | JavaScript | HTML/CSS
- Backend: FastAPI | Node.js
FitCoachAI is a mobile app powered by a multi-tool, LLM-based agent that helps intermediate-level lifters track and analyze their training progress. The app features a Simple RAG System that infers missing details from past workouts, providing additional context for more accurate responses. It also uses NL2SQL, converting natural language queries into SQL commands for logging and updating workout data. The QA system allows users to ask about their training history and receive insights based on past sessions. FitCoachAI simplifies workout tracking, helping users stay focused on their fitness goals.
I'm open to collaborating on cutting-edge AI projects, sharing ideas, or exploring opportunities. Reach out through:
📬 You can contact me via GitHub, LinkedIn, or email to discuss AI collaborations, job opportunities, or just to talk about LLMs, RAG, and AI systems.
🚀 Let’s build the future of AI — powerful, optimized, and secure.