Dean Shabi
AI engineer & data scientist building forecasting, pricing, and automation systems for energy markets.
I partner with traders, grid operators, and founders to move ideas from research into reliable production workflows - driving measurable uplifts in accuracy, risk control, and revenue.
- Experience
- 6+ years
- Forecast accuracy
- 15% → 6% nMAE
- Cost savings
- £50/MWh unlocked
About
Physics and electrical engineering roots, followed by 6+ years building production AI systems for UK/EU renewable energy trading, pricing, and grid operations. I partner directly with founders, traders, and operators to turn ambiguous market problems into measurable products.
Currently building Bloome AI, an AI-powered job matching assistant.
Selected Work
Forecasting systems, pricing engines, and optimisation models shipped with founders and cross-functional teams.
Renewcast Solar Forecasting
15% → 6% nMAE
Re-architected solar forecasting with physics-aware models and Bayesian optimisation - cut portfolio error from 15% to 6% nMAE.
High-Accuracy Forecasting Models
+57% accuracy uplift
Load, generation, and price forecasting models powering core product features and reducing balancing costs.
Portfolio Pricing Engine
Reduced portfolio VaR
Risk-based pricing optimisation enabling rapid growth while balancing risk in volatile energy markets.
Testimonials
“Dean's work has been consistently impressive - his ability to structure complex information at the right level of depth, focus on insights over raw metrics, and adapt quickly to feedback really stood out.”
Imogen
Product Manager, tem.
“Dean has a real talent for clarity - the way he presents information makes even complex, data-heavy content easy to follow. His communication style is clear, inclusive, and well considered.”
Ross
Head of Data, tem.
“Dean has been a super member of our team. His smart and thoughtful approach to complex problems set him apart. His intelligence combined with his collaborative spirit makes him an exceptional asset.”
Irene Di Martino, PhD
CEO at AmpX
“I can always rely on Dean to go the extra mile. His ability to quickly transform from a developer to a data scientist through self-learning is remarkable. His survey of time series methods became our team's foundational reference.”
Jan Bim, PhD
AI Department Lead, Datamole
Get in touch
Have a project in mind? Let's talk about how I can help with forecasting, optimisation, or automation.