10.10.2026 aktualisiert
Premiumkunde
100 % verfügbarAI/ML & Computer Vision Engineer | Full-Stack Developer | React & Django | CI/CD | Web Applications
Nürnberg, Deutschland
Weltweit
M.Sc. Computer ScienceSkills
AI AgentsRAGLLMComputer VisionMobile Application DevelopmentDjangoPythonPostgreSQLMachine LearningNode.jsTensorFlowTypeScriptReact.jsFlaskGenerative AI
Full-Stack Developer and AI/ML Solutions Lead with 7 years of experience building scalable B2B and B2C web applications and AI-powered products. My core stack is React, TypeScript and Django, and I specialize in multi-portal architectures, REST API development and end-to-end delivery from architecture decisions all the way to production.
At Elvinci.de GmbH in Nuremberg, I designed and built six role-based web portals for a high-volume B2B e-commerce platform processing thousands of transactions daily. I also implemented n8n automation workflows connecting Jira, MS Teams and PostgreSQL, and built full CI/CD pipelines using GitHub Actions and Docker.
Previously, I led the full-stack development of Kalanit Rehab, a CE-certified Class I medical device app for iOS and Android, built with React Native and TensorFlow.js for real-time AI-powered pose detection. The app was distributed exclusively by Chiesi Italia and formed the basis of my Master’s thesis at Sapienza University of Rome.
Alongside my full-stack work, I lead AI/ML projects together with a dedicated development team. This gives clients a single point of contact and technical ownership on my side, with additional delivery capacity behind it for larger projects. Recent work includes:
- An AI-assisted campaign booking portal with OCR, vision LLM and CNN-based document classification, automated pricing and Stripe payments (Flask, PostgreSQL, Google Cloud, Vertex AI)
- A multi-agent RAG assistant for analyzing large volumes of nuclear industry documentation, answering questions with source citations (LangGraph, Claude via AWS Bedrock, Pinecone, Streamlit)
- An AI medical writing pipeline that drafts M11-compliant clinical report sections using planner, researcher and writer agents with evidence from PubMed (LangChain, OpenAI, FAISS, FastAPI)
- A data pipeline that collects listings from major UK property auction platforms and flags below-market-value opportunities for investors (Scrapy, Selenium, pandas, PostgreSQL)
AI-assisted development is a natural part of how I work today. I actively use tools like Claude, ChatGPT and Gemini to move faster without sacrificing code quality, and I build automation workflows that reduce manual work across teams.
I hold an M.Sc. in Computer Science from Sapienza University of Rome.
Available for projects in full-stack development, LLM integration, RAG and multi-agent systems, and workflow automation.
Erfahrung
Berufserfahrung5 - 10 Jahre
Freelancer seit2018
Sprachen
DeutschgutEnglischverhandlungssicher
Projekthistorie
Designed and built 6 role-based web portals (Admin, Sales, Back-Office, Suppliers, Buyer Portal, Warehouse Management) with React and TypeScript for a B2B e-commerce platform. Developed scalable Django REST APIs for a high-throughput Order Management System with automated invoicing. Implemented n8n automation workflows integrating Jira, MS Teams, and PostgreSQL. Built end-to-end CI/CD pipelines using GitHub Actions and Docker. Leveraged AI-assisted development tools (Cursor, ChatGPT) for debugging, refactoring, and technical documentation.
Led the technical development and full-stack implementation of Kalanit Rehab, a CE-certified Class I medical device app for iOS and Android for back and neck rehabilitation, distributed by Chiesi Italia. Implemented a real-time full-body pose detection engine using TensorFlow.js and MediaPipe. Built the clinical dashboard as a full-stack feature using React Native and a REST API backend. Developed reusable UI component systems with React Storybook and Material UI. Project submitted and defended as a Master's thesis at Sapienza University of Rome.
Developed core features for social media applications using React, including real-time chat (WebSocket), user profiles, and activity feeds for thousands of active users. Introduced and enforced frontend coding standards, component libraries, and architecture guidelines to improve scalability and maintainability across the development team.