Data & AI Scientist · Bonn, Germany
Powering the energy transition with AI and climate insight.
A Data & AI scientist with nearly 20 years in weather, climate and energy. I lead and spearhead AI initiatives for Europe’s energy networks, preparing the grid for the net-zero transition – and I build useful AI products, too.
I bridge the gap between complex environmental data and real-world solutions, working at the intersection of weather and climate science, AI and energy.
My path ran from numerical weather prediction in Tehran to fieldwork in Antarctica, from leading R&D at New Zealand’s national weather service to spearheading AI for Europe’s energy networks in Germany.
What drives me is using artificial intelligence to optimise energy systems, deepen our understanding of a changing climate, and deliver sustainable outcomes for business and the planet.
Tailored strategies for complex data challenges – grounded in your goals, so every solution is data-driven and fits what you actually need.
Leading and spearheading AI and data science initiatives that prepare Europe’s energy networks for net zero – from smart energy services to the grid of the future.
Machine-learning weather prediction – models trained on decades of atmospheric data that now match physics-based forecasts at a fraction of the cost – and what it changes for energy: larger ensembles, site-specific calibration and probabilistic forecasts the grid can act on.
Resource assessment and forecasting for wind and solar sites, built on boundary-layer and mesoscale meteorology, high-resolution models and refined statistics.
Designing AI agents and automated workflows with large language models, custom MCP servers and low-code orchestration – from research assistants to Echo, the scheduling assistant on this site.
Agents and AI systems I have designed and run – on my own infrastructure and in the products of Agile With AI.
How I build themAn orchestrating agent that delegates to specialised sub-agents for communication, scheduling, research and content production – each with its own tools, memory and language model – and reports back through a single conversational interface.
Documents from cloud storage are ingested and embedded automatically; the agent combines semantic search with SQL over structured data and maintains a persistent memory of every conversation.
Classifies incoming requests and their sentiment, retrieves answers from a knowledge base with vector search and re-ranking, drafts replies, and routes each task to the most suitable hosted or private model.
Model Context Protocol servers that give any AI client one shared, governed toolset: web research, messaging, live data, and natural-language questions over a relational database.
MeetingEcho transcribes meetings in around 100 languages, detects the spoken language sentence by sentence for live captions and writes notes with the model of your choice – with recording and transcription kept entirely on the device.
Decomposes a question into parallel searches, condenses the sources with several language models into structured, verifiable findings, and publishes the report to a shared knowledge workspace.
Alongside my day job I run Agile With AI, a small product studio. We take everyday problems, prototype quickly with AI, test with real people – and only release what genuinely helps.
Records and transcribes meetings, dictates into any app and turns what colleagues say into language practice – all on your Mac, without a bot in the call.
Finds new public tenders for architects and engineers in their field and region, and sends only the ones that fit.
New ideas are being tested right now. They get shown when they are genuinely useful – not before.
E.ON Digital Technology · Germany
I lead and spearhead AI and data science initiatives for Europe’s energy grid, preparing networks and smart energy services for the net-zero transition.
I work with a brilliant team of data and AI scientists, engineers and analysts providing technical and scientific solutions for many aspects of energy and the energy transition.
Freelance · Germany
Technical and scientific solutions for next-generation numerical modelling, AI-based weather prediction models and cloud-based data science infrastructure, and support with environmental modelling requirements.
MetService & MetraWeather · New Zealand
Led operational and R&D work in numerical modelling, data science, satellite processing and AWS-based HPC; managed stakeholders and acted as product manager for modelling and data science.
Core expert on the World Meteorological Organization (WMO) Expert Team on Aeronautical Meteorological Science.
MetService · New Zealand
Led the numerical modelling team, advised renewable energy clients (wind and solar) on resource assessment and demand prediction, and moved an on-demand modelling system to the cloud.
MetService · New Zealand
R&D on high-resolution numerical prediction; maintained and tuned NWP models and improved post-processing systems.
University of Canterbury · New Zealand
Studied the climate of the McMurdo Dry Valleys in Antarctica (wind and energy fluxes) and modelled snow transport over Antarctic sea ice.
Helped build weather and climate prediction tools on climate-change impacts for the wine industries of New Zealand and South Africa.
Atmospheric Sciences and Meteorological Research Center · Iran
Post-processing of ensemble prediction systems with Bayesian Model Averaging (BMA).
Institute of Geophysics, University of Tehran
High-resolution numerical and statistical simulation of wind patterns and mesoscale circulations over an urban area near complex terrain, including pollution distribution.
Institute of Geophysics, University of Tehran
Climate modelling and downscaling of climate scenarios: the orographic influence of the Zagros Mountains on the climate of their leeward side.
Air University
Graduated as an air traffic controller.
Peer-reviewed work on numerical weather prediction, Antarctic climate, wind and ensemble forecasting. The full list is on Google Scholar.
How to build almost free AI agents with n8n, a free-tier cloud server and hosted language models – a practical guide for personal use and small businesses.
AI · AI agents · n8n · small business · DIY
How machine learning advances solar and wind energy – from forecasting and site selection to maintenance, grid integration and energy policy.
Renewable energy · Wind energy · Machine learning · Energy · EU
Bill Gates' How to Avoid a Climate Disaster, read by a meteorologist: 52 billion tonnes of greenhouse gases a year, and why renewables alone will not reach zero.
Climate · Climate change · Renewable energy · Book notes
I’d love to hear about it. Echo, my assistant, can pass on a message or find a time for a call – or connect with me on LinkedIn.
Iman’s AI assistant
Hi there! I’m Echo. I can schedule a meeting with Iman or pass on your message by email. How can I help?
Echo is an AI and can make mistakes. Your messages are processed as described in the privacy notice.
