Full-Stack AI Engineer with 5+ years of production experience, currently working as an AI Engineer at Oktatron. I build TypeScript/JavaScript-driven web applications and Python AI backends — including RAG pipelines, LLM integrations, and autonomous agents using LangChain and LangGraph. Java and Mendix experience available where needed; current focus is the TypeScript + Python AI stack.
Building the future AI solution in a startup — details are classified for now. :)
Enterprise CRM development on Mendix platform. Built AI-enhanced sales automation workflows using the Perplexity TypeScript SDK, integrated Apollo.io and Calendly APIs.
Admin platform development with Angular + Java Spring Boot
Development of enterprise financial/accounting web applications. Delivered a Nuxt.js + Firebase marketing site and handled NAV XML integrations for Hungarian tax authority compliance.
Legacy PHP → Angular 17 + Java Spring Boot modernization
A practical write-up on moving from low-code development into full-on AI engineering — covering system prompts, behavior rules, context engineering, codebase intelligence, and the stack that makes coding agents faster, cheaper, and more predictable.
No vector database. No embedding model. No infrastructure beyond a single serverless function. I built this on Cloudflare Workers with React Router 7 and TypeScript. The app lets you upload PDFs, ask natural language questions, and get structured answers back — all running at the edge for pennies per month.
A fully agentic AI system that autonomously plans, researches, and executes project tasks. Built with LangGraph, NVIDIA NIM, and real-world integrations: Notion, Jira, and Slack.
No vector database. No embedding model. No infrastructure beyond a single serverless function. I built this on Cloudflare Workers with React Router 7 and TypeScript. The app lets you upload PDFs, ask natural language questions, and get structured answers back — all running at the edge for pennies per month.
A fully agentic AI system that autonomously plans, researches, and executes project tasks. Built with LangGraph, NVIDIA NIM, and real-world integrations: Notion, Jira, and Slack.
Production-grade fraud detection pipeline with model monitoring and explainability.
Browser-based accessibility tool using real-time face landmark detection via MediaPipe and WebAssembly.
Google Hackathon project. AI assistant Chrome extension.
Let's connect and discuss your next project
bzddev@pm.me
For project inquiries, collaborations, or just to say hi
→Professional Network
Connect professionally and follow my updates
→Code Repository
Browse my open source projects and contributions
→Typically responds within 24 hours