Brent Positioning Framework — co-developed with Improm.ai, backtested across 14 years with 13 of them profitable, and trading live with real capital since March 2026 — is the first publicly available strategy. Company will pitch at Nexus Luxembourg 2026.
Zenpulsar, the London-based commodity intelligence company, today announced the public launch of its first systematic trading strategy, marking the company’s move from AI-generated commodity signals into AI-built, tradable investment frameworks for systematic investors. The strategy demonstrates that AI can be an effective tool for generating, testing and combining investment signals at the quality bar quant trading desks require.
The first strategy released, the Brent Positioning Framework, co-developed with partner Improm.ai, has been backtested across a fourteen-year track record, thirteen years of which were profitable. The framework has been evaluated using institutional backtesting standards, including transaction-cost assumptions and drawdown analysis. It has been trading live with real capital since March 2026 and is in production at a $100 billion utility company. A live demo is available at brent.zenpulsar.com.
The $20 trillion global commodity market is driven by the convergence of thousands of signals — news, weather, production, shipping, geopolitics, sentiment, tariff and policy. Most trading desks still transform raw data into tradable signal by hand, burning over 60% of their researchers’ time on it and still operating only partially informed. Zenpulsar was built around that bottleneck.
The Zenpulsar Agentic AI Platform — developed inside the Google AI First Accelerator (Cohort ’23) — combines two production layers. The first is live commodity intelligence: daily-scored news, weather, production and shipping events sourced from institutional sources and transformed into quant-grade signals through dashboards, an API and a multi-agent research layer. The second is the strategy layer, where multiple signals are converged by AI agents into ready-to-trade systematic frameworks. Raw data in. Tradable signal out.
Today’s launch marks the first time Zenpulsar is making its strategies — not only its underlying signals — publicly visible. It is a demonstration that AI can be deployed effectively across the full systematic-investor workflow: generating signal candidates from raw data, validating them, and combining them into frameworks robust enough to trade live with institutional capital.
“Raw data doesn’t trade. Signal does,” said Alexander Pisemskiy, co-founder and CEO of Zenpulsar. “We built an agentic platform to turn the chaos of commodity data into systematic strategies that can be traded. The Brent Positioning Framework is the first one
we’re showing publicly — backtested across fourteen years, thirteen of them profitable, and trading live with real capital since March.”
Zenpulsar will pitch the platform and the strategy in the Investment Intelligence session on the Pitch Arena stage at Nexus Luxembourg 2026, taking place 10–11 June at LuxExpo The Box.
