I am a Frontend Developer specializing in building clean, fast, and scalable web experiences with React and Next.js. Over time, my curiosity about how data flows behind the user interface led me to dive into Node.js backend development and scripting with Python.
Today, that same curiosity has pulled me into the world of Generative AI, LLMs, and AI Agents. I'm actively learning how to build workflows where code doesn't just display information, but intelligently interacts, reasons, and automates tasks. I'm on a journey to transition into an engineer who designs and deploys intelligent agentic systems.
- Frontend: React, Next.js, JavaScript (ES6+), HTML5, CSS3
- Backend & Scripting: Node.js, Python, FastAPI, SQL
- AI & Automation: LLM Orchestration, RAG (Retrieval-Augmented Generation), AI Agents, Vector Databases (ChromaDB)
- Tools & Infrastructure: Git, Docker, Postman, AWS Fundamentals
Here is a look at how my focus is evolving from full-stack applications to AI-driven automation:
- multi-tool-math-toolkit-agent A hands-on exploration of agentic workflows. Built a basic math toolkit agent capable of performing complex math calculations and autonomously searching Wikipedia for contextual answers.
- YouTube-RAG-Chatbot-API A Python FastAPI backend allowing users to chat with YouTube videos. It extracts transcripts, processes and embeds them into a local ChromaDB vector store, and uses a local LLM via Ollama to serve answers.
- spreadsheet-AI-automation A workflow automation project leveraging AI agents to parse, update, and manage spreadsheet data intelligently.
- marketing-roi-project An ETL pipeline and analytics project built with JavaScript, Python, and FastAPI to evaluate and compare the effectiveness of different marketing strategies.
- spotify_recommender A sleek, modern web application designed to curate personalized Spotify playlists based on user preferences in genres, eras, and languages.
- Advanced multi-agent orchestration frameworks.
- Optimizing local LLM inference and embedding pipelines.
- Designing self-correcting RAG systems.
- LinkedIn: in/guyonakeyboard
Always curious, always learning, and always building. ๐