Built by practitioners who lived the four-feed problem.
Helios AI was founded in Minneapolis in 2025 by a team of commodity data analysts and infrastructure engineers who spent years watching procurement desks make decisions from incomplete pictures. We built the layer that was missing.
Why Helios AI exists
The agricultural commodity intelligence market was not short on data. It was short on synthesis. Weather data sat in one tab, USDA crop reports in another, Baltic Dry in a third, and spot price history in a spreadsheet someone built two years ago. Each source was useful. None of them talked to the others.
We built Helios AI to do the synthesis work that procurement analysts were doing manually, inconsistently, and at the speed of one person checking four feeds every morning before a buying decision. The four-signal methodology emerged from studying which combinations of inputs had historically preceded the price moves that hurt procurement margins most.
We are based in Minneapolis because the Upper Midwest sits at the center of the North American grain and protein supply chain. Founded in 2025 and angel-backed, we are an early-stage team building toward a product that surfaces supply risk before it shows up in price. We are not trying to replace the data sources a desk already tracks. We are building the layer that synthesizes what those sources say separately into something actionable together.
The team
Francisco Martin-Rayo
CEO and Co-Founder
Francisco's background is in commodity data analytics, with particular focus on price signal pipelines for food and agricultural supply chains. He co-founded Helios AI after spending years observing how procurement teams consumed market data: four tabs, three formats, one analyst trying to reconcile them before a buying decision.
Sarah Blackwell
CTO and Co-Founder
Sarah works on time-series forecasting systems and agricultural data infrastructure. Her prior work spans data ingestion pipelines and API-first product development across multi-source agri data sets. At Helios AI she owns the core forecasting stack: the signal blending architecture, the probability distribution outputs, and the data delivery layer that feeds both the dashboard and the API.
Omar Fadel
Head of Data Science
Omar works on crop yield modeling and agri-climate signal integration. His focus is the translation layer between raw gridded weather products and commodity-specific yield forecasts: how to score NDVI readings, ground-station data, and USDA crop progress reports against each commodity's historical yield-price relationship. He built the signal weighting methodology at the core of the Helios AI forecast engine.
Get in touch
Questions about coverage, the methodology, or early access? Reach out directly.
- [email protected]
- +1 (612) 214-0168
- 80 South Eighth Street, Suite 900, Minneapolis MN 55402
Working with Helios AI
Send us a messageReady to put four signals to work for your procurement team?
We are taking early-access applications now. Tell us which commodities you buy and how you currently track price risk. We will take it from there.