Jordan Dimov
Energy Trading Advisor | Founder, A115
The most expensive trading technology mistake is never a bad line of code. It's a bad architecture decision made six months before anyone realises.
I'm Jordan Dimov. I advise European power and gas trading firms on platform strategy, technology architecture, and the partnerships that make both work. I also work with investors entering European energy markets — helping them evaluate trading teams, assess technology risk, and navigate a market that punishes outsiders who move without local intelligence.
Through A115, I've spent over a decade building the production systems behind some of Europe's most active trading desks — at Shell, Centrica, and Limejump. That experience shapes everything I do now — not because the code matters, but because it taught me where platform decisions go wrong and what it costs when they do.
The Problem I Keep Seeing
Trading firms building technology in-house face something I call the builder's blind spot. The team is deep in the problem, making reasonable decisions every day, and the platform looks fine — until it meets production reality. A position aggregation model that doesn't account for unit-of-measure mismatches across markets. A P&L pipeline that works at low volume but can't reconcile at scale. A settlement engine designed around one exchange that breaks the moment you add a second.
These aren't junior mistakes. They're structural design flaws that only become visible under real trading conditions — and by then, the cost is measured in months of P&L, not hours of rework.
The firms that avoid this aren't necessarily smarter or better-resourced. They bring in someone who's already seen what breaks.
How I Work
Every engagement starts with a conversation, not a proposal. I want to understand what you're building, where you think the risk is, and — more importantly — where you haven't looked yet.
Most relationships begin with architecture validation: a focused assessment of where a platform's real design risk sits. It's not an audit and it's not a code review. It's a senior practitioner asking the questions your internal team is too close to the problem to ask themselves.
From there, it depends on what you need. Some firms want ongoing strategic advisory — a trusted external perspective on platform direction, vendor decisions, and build-vs-buy trade-offs. Others need hands-on technical delivery: exchange integration, real-time P&L systems, or data pipeline architecture. And increasingly, I connect trading firms with the right capital partners and technology talent — because in European energy, knowing who to work with matters as much as knowing what to build.
Where This Comes From
Shell Energy
I built multi-layer P&L systems spanning front, middle, and back office: real-time fast P&L for the trading desk, human-estimate layers for trader judgment, and official reconciled P&L for finance. I designed unit-of-measure standardisation across TTF, THE, PEG, and NBP — solving the problem of comparing positions across markets that measure gas differently. A major focus was performance optimisation across the full stack — Python, SQL databases, Azure data lakes, Apache Pinot, and Kafka — to improve visibility, reduce latency, and give traders better quality, more actionable data as they need it. I also integrated with ETRM systems like FIS Aligne and OpenLink Endur, bridging the gap between bespoke trading tools and the enterprise platforms they have to coexist with.
Centrica
High-volume transaction processing at 100,000+ per minute, with real-time analytics on streaming infrastructure. I built complex, high-throughput data pipelines that fetched market data and fundamental data from highly diverse, heterogeneous sources — sanitising, transforming, and making it available to trading desks, machine learning teams, and risk analysts in near real-time. The kind of scale where architectural shortcuts from year one become existential problems by year three.
Limejump
EPEX and N2EX day-ahead trading automation for a fast-scaling independent generator. Exchange integration built from scratch, in production, handling real money.
Each of these taught me something different. Shell taught me that full-lifecycle complexity is where most platforms silently fail. Centrica taught me what happens when you design for today's volume instead of next year's. Limejump taught me how to move fast without cutting the corners that matter.