Generative AI comes to supply chain management

ChatGPT is now bringing efficiency and effectiveness to supply chain data. I sat down with Francisco Martin-Rayo, co-founder of Helios, which just launched Cersi, a tool that monitors and delivers insights on climate, economy, currencies and political risks, as well as delivers news articles through its database.

Here is what Martin-Rayo has to say about the intersection of ChatGPT, supply chain efficiency, and making analysts’ jobs a lot easier:

Keesa Schreane: What supply chain challenge is Cersi solving and for whom?

Francisco Martin-Rayo: We help global CPGs, food processors, and commodity trading corporations predict and mitigate agricultural supply chain risks. With increasing climate catastrophes and crop failures globally, we're helping to make sure preventable food shortages never happen again. Our artificial intelligence platform aggregates billions of data points to provide our customers with a global, real-time view into the climate, economic, and political risks affecting their suppliers, plus tracking force majeure events.

KS: Your co-founder said this tool 'will change the game for any Fortune 500 company that purchases agricultural commodities.' Can you differentiate how the tool may be used in different use cases? For example, is there a use case for institutional investors, versus bankers, versus risk officers?

FMR: So far, we're fully focused on working with companies that purchase or deliver the physical commodity (e.g., think a global CPG that needs to get mayonnaise on the shelf, or a global fast food chain that needs to get fries in your hand).  For all of them, they track their global agricultural supply chain risks with us in real-time.  This way they can track, at the earliest stages of the growing season, which suppliers are likely to have a difficult growing season so they can mitigate that risk.    

KS: Does Cersi primarily get data from large data providers, or from scraping public websites? How does generative artificial intelligence (GenAI) tech shift the data source component?

FMR: We get our data from a wide range of sources, both open-source and proprietary. Gen AI for us hasn't shifted the data source component, but rather it has given us the ability to answer custom questions our users have using natural language - in seconds, instead of weeks! For example, our latest GenAI product, Cersi, can answer, "What are the biggest climate risks to my supply chain today?" in a matter of seconds. This is transformational for companies that need to wait weeks or outsource analysis like this and will dramatically upskill procurement teams worldwide.  

KS: How does Cersi present a different value prop than already exists with current geospatial and physical climate risk start-ups and activist data providers?

FMR: Cersi is revolutionary in the agricultural supply chain space, because it empowers professionals to swiftly address questions that previously required weeks and significant financial resources. For instance, it can rapidly identify the most vulnerable farms among 150,000 in their supply chain. This is a game-changer for teams in a time-crunched industry, providing answers in seconds rather than weeks.

Cersi acts as an ever-available, highly intelligent supply chain analyst, addressing queries like the impact of El Niño on your supply chain or the optimal coffee sourcing location between Colombia and Brazil. It condenses Helios' extensive insights and tailored historical data into an intuitive chat interface accessible to all. Other products in the market require the supply chain analyst to sift through dashboards and data, rather than providing them with custom insights and responses to their specific questions. This dramatically increases the quality of decisions they can make and helps them proactively mitigate these risks, increasing their supply chain's resilience.

KS:  Do your 'dependencies' for scaling Cersi include ensuring data quality, along with financing? How are you working to mitigate data quality risks? Can the GenAI component impact data quality?

FMR: Our primary dependency on scaling Cersi is less data quality than customer interaction. The more customers we have, and the more varied their questions to Cersi, the faster and better she becomes at understanding the risks and nuances of their agricultural supply chains. It's the most powerful aspect of GenAI and what's most exciting to us is that the more you engage with a particular product the better it gets. We're lucky that we just announced our $2M pre-seed round and are excited to use these resources to triple down on our extraordinary technology.  

KS: What do you hope clients will say about Cersi in the next 1-2 years?

FMR: Within a year our goal is that procurement managers globally think of Cersi as the best supply chain tool they've ever had.