Water is the world’s most critical and least understood resource. Powered by advances in AI and satellite observation, Tova (Techstars 2025) brings water risk expertise in-house — giving water utilities and water-intensive industries the power to understand, forecast, and act on water risk with unprecedented accuracy. On demand. Any facility, anywhere. 10x cheaper and 90% faster than traditional approaches.
Emilie Vallauri, Global Program Manager at Techstars, sat down with Alexa Bruce, CEO and co-founder, to find out how.
Emilie Vallauri: To start out, could you introduce yourself and tell us a bit about your background?
Alexa Bruce: I’m a water resources engineer — civil engineer by training. I’ve been focused on water my whole adult life; it’s always been a passion. I started at Arup, one of the big global engineering consultancies, where I led a multi-city water resilience programme spanning Miami, Cape Town, Amman, and Mexico City. Then I moved to the client side — at the UK’s National Infrastructure Commission, I was part of the team that produced the country’s first long-term integrated national infrastructure plan. So I went from building models to being the person who had to use them to make decisions.
That switch in perspective changed everything. The best available water science — the kind that should be informing critical decisions about supply, infrastructure, and climate adaptation — has always struggled to reach the people who need it at the speed and scale that those decisions demand. I recognised that the expertise needed to change that lived in academia, so that’s where I went.
Emilie: And who are the other members of your founding team? How did you meet?
Alexa: I did my PhD at UMass Amherst with Casey Brown, my co-founder. Casey is our Chief Science Officer. He’s spent 25 years at the frontier of hydroclimatic scientific discovery — his research has shaped how utilities, governments and banks around the world think about water risk. He’s a Provost Professor at UMass Amherst. We met about eight years ago at World Water Week in Stockholm, worked together for a couple of years whilst I was still at Arup, and he eventually recruited me to do my PhD with him.
Beyond the two of us, we’ve built an exceptional team of seven spanning hydroclimatology, water resources engineering, AI, product, and finance. Several came out of our research groups at UMass and Cornell. The depth of shared context across the team is something you just can’t hire for overnight.
Emilie: How did the idea for Tova start? Was there an “aha” moment?
Alexa: It was more of a gradual culmination of our life’s work. It wasn’t a case of “let’s start a startup and brainstorm ideas.” It was Casey’s decades of scientific leadership crystallising into a product, combined with my recognition that the way water resource analysis gets done today is fundamentally broken. A single basin study can take months and cost hundreds of thousands of dollars, and the moment you have a new location, a new question, you start again.
The real cost isn’t just bad decisions — it’s the decisions that never get made because the data arrived too late or was too expensive. A startup was the only vehicle that could deliver at the cadence and quality required.
Emilie: Simply put, what does Tova do and why is it so important?
Alexa: Water touches everything — public health, food security, energy, insurance, infrastructure. When a utility underestimates drought, people can lose access to clean water. When infrastructure is planned with outdated data, the consequences can last decades. Yet the vast majority of river basins globally lack the data needed for reliable planning.
We built Tova to close that gap. RiverCloud is the engine — our physics-informed AI model that simulates how entire basins behave: surface water, groundwater, snow, reservoirs, withdrawals, the lot. Pisces is the interface — an AI agent embedded in customer workflows. You can ask it questions, run scenarios, compare interventions, and get decision-ready outputs. It’s like having a water expert on your team, available on demand.
The goal is simple: make water risk visible before it becomes a crisis, and give people the tools to act on it.
Emilie: How is deep learning used on the platform, and what data are you training on?
Alexa: The IP is really in our models. RiverCloud leverages deep learning but constrains it according to known physics, so we can explain exactly what’s driving model behaviour. That’s essential when you’re working with utilities and infrastructure operators who need to trust the outputs.
A key data input is the SWOT satellite, launched in 2024 by NASA and CNES. For the first time, we can observe water surface elevation, slope, and extent across rivers, lakes, and reservoirs globally. The data is technically public, but turning raw satellite radar into reliable hydrologic inputs requires years of specialised expertise — only a handful of research groups in the world can do it, and our team includes people who’ve been at the forefront for years.
Emilie: And who are your primary clients?
Alexa: We currently have a contract with East Bay Municipal Utility District in California. Utilities are the natural starting point because they manage water systems in real time and need the most precise understanding of supply risk.
Beyond utilities, food and beverage companies are a major target — facilities that depend on surface water in regions where nobody could previously provide reliable models. Tova gives them forward visibility where none existed before.
We also see a massive opportunity in parametric insurance — products that trigger automatic payouts when, say, river levels drop below a threshold. That only works with reliable, near-real-time hydrologic data. With ours, entirely new forms of water-risk insurance become possible.
Emilie: What sets Tova apart from traditional engineering consultants or climate dashboards?
Alexa: The status quo is bespoke modelling — high quality, but it takes months, costs hundreds of thousands/ millions, and the moment you have a new question, you start again. Risk dashboards sit on top of coarse global datasets that are useful for disclosure but not reliable enough for material decisions. And there are one or two emerging players applying deep learning to streamflow forecasting, but this is really only just emerging — our MVP results are literally unprecedented, even by academic standards.
Tova is a different category. We combine the fidelity of bespoke modelling with the speed and scale of AI, delivered continuously. Ten times cheaper and ninety percent faster.
Emilie: There are so many things you have achieved since you founded Tova. What has been your proudest milestone so far?
Alexa: The thing I’m proudest of is the team. What we’ve shipped — a physics-informed deep learning model that outperforms industry benchmarks and delivers literally unprecedented results on reservoir operations — that’s the product of a group of people who are absolutely world class and steaming ahead.
But the moment that really landed has been these last few weeks, presenting our MVP to customers. You could see their eyes light up — “this is Harry Potter stuff.” When the people who know these systems best get excited about your work, you know you’re on the right track.
Emilie: What advice would you give to aspiring founders?
Alexa: You have to really, truly care about and be all-in on the mission. Beyond that, you must build up your stress tolerance. As a mum of two — including a five-month-old — I’ve found that parenthood has given me a great sense of perspective on what is actually important. If you feel every “dip” in the startup journey too acutely, it’s very hard to keep going.
Emilie: You were part of the Techstars London cohort of 2024. Why did you participate in Techstars, and how did it help you and Tova on your journey?
Alexa: Entering the programme, I wasn’t entirely sure what to expect, but it gave us so much! Coming from an academic and engineering background, I hadn’t engaged with the “dating dynamic” of fundraising before. Techstars taught me the rules of the game. Georgie (my MD) is an absolute rockstar and continues to be so supportive in a really hands-on way.
Techstars’ global mentor network has also proven incredibly valuable for us. I still meet with some of my mentors every other week, so that relationship has persisted well beyond the programme.
