From the Supply Change x Helios AI webinar on real-time foresight for procurement leaders

The new normal: climate shocks + geopolitics + margin pressure

The starting point for this conversation is simple and uncomfortable: food supply chains are under pressure like never before.

Helios AI’s Co-Founder & CEO, Francisco Martin-Rayo, and Supply Change Capital’s Co-Founder & General Partner, Shayna Harris, walked through what many procurement leaders are living every day: climate volatility, geopolitical shocks, and market disruptions are colliding faster than teams can react. Think of cocoa grown in Ghana and Côte d’Ivoire facing extreme heat and drought, shipping routes being rerouted due to conflict, and contracts becoming meaningless when spot prices double.

For a category manager or head of procurement, that volatility shows up as a constant question:

“Will I actually get the volume I contracted — and at what price?”

When a contract is signed at $6,000/ton and the market moves to $12,000/ton, relationships and spreadsheets alone can’t close that gap. Teams need a way to understand where prices are likely to land and how supply risk will evolve months before it hits the balance sheet.

“So with all this volatility — climate shocks, geopolitical shifts, and wild price swings — how is a procurement team actually supposed to keep up? How does Helios help?”
Watch Francisco Martin-Rayo break it down in this excerpt from our webinar.

Why dashboards aren’t enough anymore

Most organizations already use climate dashboards, BI reports, or pricing trackers. The problem? They mostly tell you what has already happened.

They struggle with:

• Speed: They aren’t built to ingest fast-moving climate, price, and news signals.

• Context: They show slices of data, not how those signals interact.

• Usability: As data grows, dashboards often become cluttered and harder to navigate.

In the webinar, Francisco highlighted how teams use Helios AI differently: some prefer weekly emailed reports, others want direct data feeds into Databricks or internal models, and some work inside the Helios AI dashboard and chat interface every day.

The goal isn’t to replace dashboards—it’s to augment them with a live analytical co-pilot that can answer nuanced questions, run custom analysis, and surface risks and opportunities without hunting for the right chart.

Horizontal vs vertical AI: why generic models fall short

One of the core themes in the conversation was the distinction between horizontal AI and vertical AI:

• Horizontal AI: general-purpose tools like ChatGPT or Claude that are great for broad questions, writing support, and everyday reasoning.

• Vertical AI: domain-specific systems built on proprietary data, industry workflows, and tailored evaluation.

Francisco illustrated this with a deceptively simple question about banana prices and recent tariffs. A general LLM pulls from generic internet content and returns a plausible but shallow answer: “Yes, tariffs are the main driver.”

Helios Horizon, by contrast, decomposes the question:

• How have banana prices moved over time?

• What does climate risk look like in major growing regions?

• What do news and trade flows suggest?

The result was a more accurate answer: tariffs matter, but underlying climate stress and disease in key production regions are major drivers too. That’s the power of vertical AI: it doesn’t just assemble text—it does structured analysis on the right data and shows its work.

Inside Helios Horizon: an AI co-pilot built for food supply chains

Helios Horizon is the AI co-pilot built on top of Helios AI’s climate and price forecasting infrastructure—designed specifically for food procurement and supply chain teams.

It pulls from:

• Billions of signals—climate data, Helios AI’s commodity-specific price models, USDA and industry reports, and 250,000+ news sources updated every 15 minutes.

• A multi-agent system—climate, price, news, and availability. Sub-agents work together under a supervisory agent to ensure accuracy.

• Transparent performance metrics—users can see how well forecasts performed historically for crops like blueberries in Peru or tomatoes in Spain.

• Full citations—sourcing back to USDA, academic papers, and Helios’ own platform.

In practice, it feels like having a senior analyst on call:
Ask, “What’s the outlook for Roma and round tomatoes next season?” or “Where are my biggest climate and availability risks across my top 20 commodities?” and get a multi-layered, sourced answer in minutes.

Real-world ROI: from mandarins to Brazilian oranges

The conversation grounded the tech in real stories:

• Libby’s example—timing the mandarin market
  Helios worked with Libby’s, one of the largest buyers of pineapples and mandarins in the U.S. Helios’ price forecasts showed mandarin prices were set to climb sharply. Libby’s team validated the data with their on-the-ground intel, then bought more volume earlier than they normally would. When prices later rose by ~15%, they had already secured supply—generating meaningful savings in a low-margin category.

• Brazilian oranges & Peruvian mangoes—seeing the shock first
  Helios identified stress in Brazil’s orange crop and in Peru’s mango production months before it showed up in mainstream reports. Customers using Helios could act on those early signals, adjusting sourcing strategies or finding alternative origins before the rest of the market started scrambling.

This is where predictive AI stops being abstract. When you’re negotiating multi-year potato contracts or managing specialty crops with thin data and high volatility, seeing the risk 6–12 months ahead can be the difference between a controlled plan and a crisis.

From reactive to proactive: how roles are changing

One of the most interesting parts of the webinar was how different roles use Helios AI:

• Procurement analysts
  Use the platform daily to pull price forecasts, climate risk, and news into reports and risk registers. Helios makes it feasible to cover all key commodities and suppliers, not just the top few.

• Category managers & procurement leaders
  Check in weekly or biweekly to see which categories and regions are flashing red, what to prioritize in upcoming negotiations, and where to adjust strategy for the next 3–6 months.

• Chief Procurement Officers/heads of ag portfolios
  Use Helios for strategic horizon scanning:

• Will we still be able to source the same volume and quality from key regions in 5–10 years?

• Do we need to rethink where we’ve built plants or processing facilities?

• What does climate risk look like across a $4B agriculture loan portfolio?

Predictive AI doesn’t replace procurement talent—it makes it possible for lean teams (and almost all procurement teams are lean) to operate with the kind of foresight that used to require army-sized analyst groups.

Democratizing foresight: Walmart to SMBs

Helios AI works with some of the world’s largest food companies, including Walmart, where Helios became the first software company to win their Open Call program.

But the same infrastructure is intentionally priced and packaged for mid-market and SMB manufacturers, processors, and brands.

Foresight shouldn’t be limited to Fortune 50 budgets. With Helios AI, a $100M family-owned food business can access the same level of intelligence as a global retailer.

What’s next—and how to explore it for your categories

Francisco outlined a future where multi-agent AI automatically scans for second- and third-order effects:

“A poor raspberry harvest pushes buyers into another region… which drains supply for a different product… which begins to shift prices months ahead.”

We’re only at the start of this trajectory, but predictive, domain-specific AI is already separating teams that anticipate disruption from those that react to it.

If you’d like to see what this looks like for your own business, book a live demo or proof of concept with the Helios AI team: https://www.helios.sc/requestademo.