Projects that I have built and contributed to.

From enterprise RAG assistants and agentic AI systems to full-stack ERP platforms — these are the projects I am most proud of.

  • Enterprise RAG Assistant (Konecranes)

    Internal enterprise knowledge assistant over heterogeneous documents (XML, PDF, XLSX), with multiple RAG variants (Bedrock-native, LangChain, LangGraph), a shared conversational UI, multi-agent orchestration, and a feedback-based evaluation pipeline. Built with AWS Bedrock, FastAPI, Next.js, PostgreSQL, and AWS Amplify/ECS/ECR/RDS.

  • Health Information RAG Chatbot

    MSc mini project at the University of Jyväskylä — a production-grade domain-specific RAG chatbot built by scraping a public health information website, processing content into a ChromaDB vector store, and integrating an LLM for conversational health information retrieval. Full pipeline: web scraping → document parsing → vector ingestion → conversational retrieval interface.

  • United Used Cars Website and MIS

    Full website and management information system for a used car trading company. Built with Laravel, Next.js, Material UI, and MySQL.

    unitedtradingcars.com

  • PGL ERP System

    ERP system for Peace Global Logistics covering logistics, shipments, HR, and inventory modules. Built with NestJS, React, Next.js, and PostgreSQL.

    pglsystem.com

  • TGL Supplies System

    High-performance website and management information system built with Node.js and Express.js for the backend, React.js for the frontend, and PostgreSQL for the database.

    tglsupplies.com

  • Charity Management System

    Charity management system built for an Afghanistan-based non-profit organization.

    github.com

  • TeebAlhoor ERP System

    Complex ERP system for TeebAlhoor, a UAE-based trading company, built with Laravel, Nuxt.js, and PostgreSQL.

    github.com

  • Peace Global Tech Website

    Department website for the tech division of Peace Global Logistics.

    peaceglobaltech.com