Applied AI for energy markets.

Independent validation for trading desks. Customer analytics for energy suppliers. AI roadmaps for grid operators. Founder-led, grounded in hands-on experience running German energy market processes.

Trade. Sell. Deliver.

Energy is one value chain with three different businesses inside it: the desks that trade power, the suppliers that sell it, and the operators that deliver it. Generic AI advice fails all three, because the value hides in the specifics — merit-order dynamics and regime shifts on the desk, churn and tariff economics in retail, connection queues, redispatch, and market communication on the grid.

EnerCom AI works where energy and AI actually meet. We build and validate the models — not just the slideware.

What we do

Three productized engagements.

Fixed fee, fixed scope, senior delivery, and a concrete artifact at the end of every one.

Independent Strategy Validation

For algorithmic and quantitative energy trading firms.

“Our research team can’t independently check its own strategies.”

Independent backtesting and shadow trading of your strategies on out-of-sample market data — including data-leakage and regime-robustness checks — and a written validation assessment with clear keep, fix, or kill findings.

Two-week fixed-fee sprint

Commercial AI Blueprint

For commercial and marketing leaders at energy suppliers, retailers, and Stadtwerke.

“We’ve run GenAI pilots for a year. They burn budget and tokens, but nothing shows up in the P&L.”

An audit of your current AI and analytics initiatives with clear kill, keep, or scale calls — plus the two or three highest-value use cases across churn, segmentation, tariff, and campaign analytics, each with a value hypothesis, data requirements, and a 90-day pilot plan.

Two-day executive session + two-week blueprint sprint · fixed fee

Grid AI Roadmap

For network operations, asset, and innovation leadership at DSOs and TSOs.

“Everyone tells us to ‘do AI’ — but we run regulated processes on legacy systems, and we can’t afford pilots that go nowhere.”

Value-stream mapping across your core processes — connection requests, congestion management, asset maintenance, market communication — a use-case portfolio scored on value, feasibility, and regulatory fit, and two to three pilot-ready specifications.

Two-week fixed-fee sprint
Approach

How an engagement runs

1

Intro call — 30 minutes. You bring a concrete problem. We tell you honestly whether we are the right fit, and what it would take.

2

Fixed-fee sprint. Fixed scope, fixed price, senior delivery, and a concrete artifact at the end. No land-and-expand armies.

3

Scale what works. Pilot delivery, model build-out, or ongoing advisory — only where the sprint proved the value.

Every engagement is delivered by the founder, with specialist data scientists brought in as the work requires.

Who you’ll work with
Oleksii Leshchenko, founder of EnerCom AI

Oleksii Leshchenko

Founder, EnerCom AI

Oleksii Leshchenko founded EnerCom AI at the intersection of three careers that rarely meet in one person. A decade leading data-driven strategy consulting as Head of Data Consulting at Gorshenin Group — 155+ delivered projects, client portfolio doubled — followed by marketing-performance analytics at Hilti. Then a move inside the German energy market: running balancing group settlement under MaBiS in the Marktkommunikation 2.0 framework at a distribution network operator, cutting settlement cycle time by 20%. Most recently, formal retraining in data science and AI — IHK-certified, 960 hours — with a capstone forecasting German day-ahead power prices to an RMSE of €7.96/MWh. He holds a Master in Management in Strategy & Analytics from IE Business School and works in English, German, Ukrainian, and Russian.

Trade

Day-ahead price forecasting: RMSE €7.96/MWh

Sell

155+ consulting projects delivered

Deliver

MaBiS settlement run in production

LinkedIn →

If it touches the desk, the P&L, or the grid — let’s talk.

Thirty minutes, no deck, no pitch. Bring a concrete problem; leave with an honest read on whether AI can solve it — and what it would take.