AI Engineer at IBM where I architect production-grade Generative AI systems for IBM customers and business partners.
Beyond my work at IBM, I maintain Ad Astra AI—a technical blog documenting production-ready AI engineering projects, sharing deep-dives on building, deploying, and scaling real-world AI systems. My flagship project, SentiWiki AI, is an end-to-end agentic RAG system deployed on AWS ECS for querying ESA Copernicus Sentinel mission documentation
Previously, one year as a Solutions Architect at Huawei gave me deep expertise in cloud architectures, which I've further validated through AWS certifications (ML Specialty, Solutions Architect Associate, AI Practitioner) and Databricks Data Engineer certification.
Currently also an Associate Professor in Machine Learning at UAX University, where I designed and deliver comprehensive curricula for Machine Learning I & Machine Learning II courses to 20+ undergraduate students in the AI & Computation degree program.
AI Engineer @ IBM
Learn More
ML Professor @ UAX
Learn MoreProjects
🏎️ F1 AI Race Engineer
An AI-assisted replay for finished Formula 1 weekends: interactive track maps, standings, tyre and pit strategy timelines, weather overlays, race-control messages, and short natural-language takeaways so you can see how each stint and decision shaped the race—not just who crossed the line first.
Built to turn timing data and session context into something you can explore end-to-end after the chequered flag, with emphasis on strategy, consistency, and what actually moved the result.
🛰️ SentiWiki AI
An agentic RAG stack for Earth observation: Copernicus Sentinel mission docs, product guides, and terminology are indexed for retrieval, while orchestrated agents plan multi-step queries, ground answers in cited passages, and walk through EO concepts—so you get accurate, explainable responses without hunting across PDFs and portals.
Designed as an end-to-end demo of production-minded RAG—structured ingestion, retrieval with guardrails, and agent loops that favour clarity and traceability over generic chat.
Others
Technologies I use to build production AI systems