Find some of my recent projects below.
Implemented a high-performance data pipeline capturing 30+ synchronized camera, sensory, and telemetry streams of the Peer3000 during operations to enable human-like learning and autonomous intelligence.
Integrated Google's ADK framework with FUNKE's AI Platform to implement agentic use cases in production for 2,000+ active daily users.
Strategic AI consulting and development of a multi-agent AI platform using LangGraph and LangSmith, with cloud deployment and production-ready agentic use cases for a variety of clients.
Implemented an optimized RAG system using semantic search to transform siloed documentation into actionable insights. This architecture achieved a 50% improvement in search accuracy across internal company data.
Designed and built a cloud-native data platform on Google Cloud that powers real-time algorithmic energy trading. The system processes live market data, weather feeds, and grid signals to execute automated trading strategies.
Led the development of intelligent document automation for banking and insurance clients, automating classification, entity extraction, and scanning. Achieved up to 80% faster processing.
Built a cloud-native lottery slip recognition system replacing legacy scanner terminals with tablet devices. Processing 40,000+ slips monthly at >99% accuracy with zero reported issues over 20+ months in production.
Contributed to an AutoML platform empowering biomedical professionals to independently design, train, and deploy deep learning models for cell segmentation and classification without writing code.
Empowering collaborative robots to think and learn like humans 
AI-Agents Platform with Google ADK for FUNKE Mediengruppe 
Development of an AI Platform for BBN AI Studio 
Semantic Search & RAG System for Intelligent Access to Company Knowledge 
Cloud-Native Data Platform for Real-Time Algorithmic Energy Trading 
AI-Powered Cloud Solution for Digitizing Lottery Retail 
AIxCell – Democratizing Deep Learning for Biomedical Image Analysis