An AI Startup Takes Us Inside the Business of Climate Resilience

Climate change demands climate action. Add artificial intelligence to the problem-solving portfolio, and technology is helping businesses adapt to an increasingly hotter planet. In industry lingo, this is called climate resilience.
To understand this emerging field, Wharton Global Youth explored the intersection of climate change, AI, and business resilience with a startup company among the 2024 innovators selected for the Cypher Accelerator at Wharton’s Stevens Center for Innovation in Finance.
Eoliann, founded in Italy in 2022, uses machine learning and earth observation data to quantify climate risks for businesses and financial institutions. Eoliann’s tagline: Building climate resilience from the sky.
We caught up with Federico D’Albenzio, Eoliann’s business developer, to learn about the company’s software, which provides detailed, quantitative analyses of exposure to climate risks like floods and hurricanes, and how AI supports that work.
Federico, what are three things we need to understand about the business of climate resilience?
Feeling exposed. As temperatures rise, businesses are adapting to survive. For example, banks need to assess the exposure of their credit portfolios to climate risk to know if it is safe to loan money to certain clients. “Climate risk is the potential exposure of a physical asset (i.e., industrial facilities, residential buildings, agricultural fields) to extreme events, such as floods, wildfires and hurricanes,” says D’Albenzio. “We can still save the planet, but we have to adapt to it. That means we need to understand how these phenomena will change our lives. We need to be prepared to mitigate the risks.”
Data-driven decisions. Eoliann provides quantitative data on the probability, intensity and vulnerability of assets to floods and other climate-related threats. Banks, for example, can then make informed decisions about providing loans (or not, if the risk is too high) and encourage clients to take actions that will mitigate their climate risk. Eoliann uses AI to process the huge volumes of data it receives from three different satellite constellations; data that is constantly updating information about the planet. “Thanks to AI, we can process this data…and we can simplify the methodologies,” notes D’Albenzio. “AI makes it possible to use data to create new models that are much more efficient compared to statistical modeling…Our founders are very technical people who studied a lot of math and physics. Then they said, ‘How can we model our reality in numbers; in technology that has an impact?’ If you have the tools to understand the reality, and you know what you want to get in the end, AI becomes an enabler to get you there.” The innovation: detailed, numbers-driven risk assessments.
Unsung climate risks. As Eoliann grows and raises funds to scale its solutions across Europe and the U.S., it is developing a deep expertise in climate-related exposures, even those that don’t get as much airtime as high-profile hurricanes and wildfires. “Droughts are climate risks [that get less attention], observes D’Albenzio. “They are potentially so impactful because they affect every part of the value chain in the end, from energy generation, to the production of agricultural crops, to water accessibility. It’s crazily important in my point of view, and right now, people are not taking droughts into account.”
Intrigued to learn more about climate risk? Tune into a recent episode of Wharton’s Ripple Effect podcast featuring Witold Henisz, vice dean and faculty director of Wharton’s ESG Initiative, who talks about “Why Climate Risk is Financial Risk.”
What is climate resilience?
How does Eoliann build climate resilience from the sky?
Federico D’Albenzio says, “We can still save the planet, but we have to adapt to it. That means we need to understand how these phenomena will change our lives.” What does he mean by this? How are you adapting to changes brought about by climate change? Share your story in the comment section of this article.
Hero Image: Ales Krivec, Unsplash+
This article is helpful and packed with valuable information! It really nice my understanding on the topic. Thanks
This article does a wonderful job of outlining how AI startups like Eoliann are moving climate resilience from being a response mechanism to an ongoing, data-driven practice. Through the use of machine learning interpreting satellite imagery, Eoliann allows banks, insurers, and governments to quantify—and price—climate risk correctly.
1. Integrating climate risk into financial decision-making
Eoliann’s framework develops climate resilience by shedding light on latent hazards—floods, droughts, wildfires—before they become monetary loss. It brings to mind my background in AWS and Python analytics, where forecast models analyzed SMS data in underserved populations to predict financial hardship. In each instance, early warnings become strategic handles. Banks employing Eoliann can proactively tweak credit terms or demand climate-resilient plans—transitioning from reactive lending to risk-conscious stewardship.
2. Democratizing climate intelligence with AI
The article explains how AI condenses satellite information into valuable insights—a step ahead of ancient statistical models. In my data work with nonprofits, I built low-cost dashboards to track campaign outreach in real time. Eoliann is doing the same for climate: condensing massive datasets into concise risk scores. This kind of technology translates complexity into clarity—something that human-centered AI does instinctively.
3. Raising awareness about neglected climate risks
I was particularly struck by their emphasis on drought analytics—a less-spoken-about but disastrous danger. Droughts disrupt entire agriculture value chains. Similarly, at Bhoomika Trust during my internship, minor dips in donor engagement frequently preceded campaign burnout—and watching for those signals in advance enabled us to pivot. Eoliann’s sensitivity to nuanced climate data speaks to this—we need to deal with less-wild but systematic risks before they develop into crises.
Where I’d like the conversation to go next
• Integration with insurance products: Are banks or insurers using Eoliann’s risk scores to adjust premiums or tailor micro-insurance for vulnerable regions?
• Ecosystem approach: Could Eoliann partner with fintech and aid platforms, automatically triggering micro-grants or credit extensions for farmers, businesses, or communities flagged as at high risk?
• Transparency and equity: How does Eoliann prevent its models from perpetuating geographic bias—e.g., favoring wealthy regions with more data—while ensuring equitable outcomes?
Climate resilience for societal good demands more than tech—it requires orchestration: AI analytics, financial tools, policy frameworks, and social inclusion. Eoliann is making smart progress on the first two pillars; I’m curious how they’re preparing for the rest.
Thanks, Diana, for bringing global light to startups at the helm of this movement. Saving the world will indeed not be achieved in laboratories in a vacuum—it will take AI-driven vision ingrained in fiscal systems, social movements, and worldwide policy. That’s the resiliency of the future—and next-gen innovators are on the lead.
Hi Hrithik!
I enjoyed reading your comment on this paper. You connected so many different ideas. From your experience with nonprofits to your work in data and analytics, you tied them back to Eoliann in a way that made the article feel much more significant. Your point about shifting from reactive lending to proactive planning really stood out to me. I hadn’t thought about how banks could actually change credit terms based on climate risk predictions. However, it makes a lot more sense after reading your comment. It is so cool to think that AI isn’t just helping us respond to disasters after they happen, and instead, it is helping us prevent them from happening in the first place
I also really liked your opinion about droughts. You are completely right about how droughts don’t always grab much attention like floods or fires, but they can be just as damaging, especially to farmers and the agricultural community. Your example about donor engagement and campaign burnout during your nonprofit work was an amazing comparison. It shows that spotting small warning signs early is a skill that can make a huge difference. I believe that Eoliann’s work with satellite data and machine learning could become a model for how we approach other slow-building problems in society too.
I thought the way you talked about simplifying big data through AI was really clear and powerful. Turning satellite imagery into quick scores is something that could help people make better decisions. It is similar to what you did with your nonprofits, taking something complicated and turning it into something useful. Eoliann seems to be building the bridge between science and real-world actions.
You also did a great job highlighting the importance of fairness. If tools like this are going to help everyone, then the data and systems need to work for everyone too. Your point about making sure models don’t favor places with more data is something I had never thought about. You made it clear that climate resilience is about more than just technology. It is about responsibility, inclusion, and making sure no one is left out.
Hrithik, I truly learned a lot from your comment, and it made me think harder about what smart, ethical AI should look like. Thank you so much for this comment. I randomly stumbled upon this and it clearly was for good reason.
Thank you for your thoughtful comments — they truly helped me see the issue from a new perspective.
You described droughts as a “less-wild but systematic” danger, and I completely agree. We often hear about dramatic events like floods or hurricanes, but slow-moving crises like drought quietly reshape ecosystems and communities over time, often without immediate notice. This reminded me of what I learned in genetics — some mutations may not be anything in their initial manifestation, but eventually result in extreme alterations in the functioning of an organism. Similarly, global warming is an agency of gradual but gigantic power, shaping our world in tiny, incremental increments.
I also really enjoyed your idea of combining Eoliann’s AI with fintech and micro-insurance. It’s an excellent tactic to provide farmers and small businesses with timely relief, especially those who are unable to wait for regular relief. If AI can disburse funds independently and equally, it changes the game from mere survival to collective involvement in mitigating climatic threats. This raises an interesting question: how do such AI systems balance automation with human intervention to avoid unnecessary spillover, such as exclusion or discrimination?
That leads me to another concern — fairness and inclusion. Poor or rural communities usually have limited access to technology and dependable data. Would it be feasible to create AI systems that incorporate local, grassroots-based inputs — like observations, reports, or even indigenous wisdom — to augment these data deficits? This kind of participatory design would not merely add precision but create trust and ownership among users, making technology less of an outside importation and more of a collaborative tool.
Building on this, I want to challenge the ethical considerations of applying AI to marginalized groups. How do we ensure transparency in how decisions are made, data is used, and assistance is allocated? Without intentional design, these systems can entrench current inequalities or create new dependency relationships. Perhaps integrating human-centered approaches and continuous feedback cycles from the community can permit AI to evolve responsively.
Your response made me realize that building climate resilience tools isn’t just about technological innovation; it’s about creating accessible, trustworthy, and contextually aware solutions that empower all stakeholders. Thanks again for your insightful feedback — it inspired me to revisit the complex intersection of AI, climate, and equity with a more critical and hopeful lens.
This article and Eoliann’s role in climate resilience got me thinking about another potential application of AI related to environmental sustainability. Rising and increasingly unpredictable temperature swings, changing tides, and pollution have made parts of the Potomac River near my home often unswimmable, and unhealthy for many species of fish. I am part of a group that tests the water weekly. We report on turbidity, bacteria levels, temperature, and PH level. We also note tides, time of day, air temperature weather, and recent rainfall. Similar to how Eoliann predicts climate related vulnerabilities for businesses, I would like to explore building a program that uses all of the data that we collect to predict when the water might be safe for swimming, even during weeks we don’t perform out testing. Federico D’Albenzio says, “We can still save the planet, but we have to adapt to it. That means we need to understand how these phenomena will change our lives.” This made sense to me. Part of testing the river is to help identify sources of pollution to help keep the river as clean as possible. But it is also to protect the people and pets that swim or boat in the river by letting them know that at times, the river is no longer safe to swim in. As the data starts to tell a different story we communicate it. This is what I liked most about what Eoliann is accomplishing.
Landon, I particularly appreciate how you’ve connected Eoliann’s climate smarts with your in-field water-quality monitoring. That’s exactly the kind of out-of-the-box intersectoral thinking that this discipline needs.
Your idea—using AI to forecast safe swimming days—jives with the predictive model approach Eoliann uses for flood and drought forecasting. Have you thought about integrating your water-quality monitoring with real-time public alerts? For example, integrating your turbidity and bacteria monitoring with weather and tide information into an ongoing mobile alert system might help make aquatic life and swimming safer.
In my Python/AWS experience, I helped build dashboards that merge live donor metrics and campaign performance. A similar tool—merging pollution levels and ecological metrics—could trigger predictive alerts like “Unsafe conditions likely tomorrow morning.” That sort of user-facing app bridges data and actionable insight.
And another: are you considering collaborations with local authorities or local NGOs in data-sharing or co-financing? Such collaborations could enable you to standardize testing regimes and scale your model out to other rivers or urban waterways.
Finally, might explainability of machine learning models, or at least those for public trust, be studied? Citizens would be more likely to believe and act upon an alert if they understand why it’s being sent—e.g., “Bacteria levels over threshold X due to combined storm runoff.”
What frameworks or tools are you using in your water-quality data pipeline? And how are you addressing sensor calibration and data consistency across the dataset?
You’ve tapped into a powerful opportunity—turning environmental monitoring into community impact through tech. I’d love to hear more about how you’re building and testing this system!