Most AI experiments in investing and banking aren’t built on data good enough to trust with high-stakes calls. For many institutions, “adopting AI” just means prompting a chatbot with no institutional memory, no grounding in the company’s own data, and no accountability for what it says. That cuts admin. It doesn’t change what’s possible.
Henrik Landgren built data infrastructure at Spotify, then at an early-stage VC firm, before founding Gilion to close the gap he kept seeing: investors making capital decisions with almost no data behind them. Gilion replaces gut feeling with hard, well-analysed data.
In this week’s Five Minutes With…, Fintechly speaks to Landgren about why the real AI problem is a data problem, what investors still misunderstand about handing over control to AI, why good businesses go unfunded when their finances are invisible to investors, and how AI could reshape attitudes to risk.
Can you tell us about yourself and what brought you to this point in your career?
The north star of my career is figuring out how to put data to better use. First for the companies I worked for, then for investors, and now for both sides of the funding process.
I led Spotify’s data team from 2010, right as the industry was working out how to gather and utilise user experience data. Much of the data collection that no one had thought to do up until this point was now possible as smartphones and apps proliferated. We grew the company past 20 million subscribers by putting all that data to work.
I brought that passion for number-crunching to investing, where I built an early AI-powered data system at my first VC firm. It helped us track down companies worth investing in by using granular information like usage insights, so we could get to promising startups before other investors did. We were one of the first movers in Europe, and that gave us an edge.
Investing showed me firsthand how little data infrastructure is currently in place for financial decision-makers, and it’s what made me build Gilion. Now, I am helping investors allocate capital based on hard, well-analysed data, rather than the gut feeling VCs so heavily rely on.
What problem or opportunity are you most focused on right now?
I think we need more industry commitment to getting AI solutions to a point where they can actually be trusted long-term with high-stakes decisions. Every bank and fund has experimented with implementing AI into their workflow, but progress often stops at copy-pasting prompts into a chat tool that has no institutional memory, no grounding in their own company history or industry context, and most importantly, no accountability for what it tells you. It may improve productivity or reduce admin, but it’s not pushing the innovation frontier outwards. Data changed what was possible for companies and investors; AI should be able to do the same.
We tested our product for four years on our own money before offering it to other people, and the most important thing we learned in the process is that there is a world of difference between having AI as part of your strategy because you have to, and being an AI-native business.
What do you think deserves more attention than it is getting in your part of the industry?
The data problem sitting underneath the AI hype. Everyone knows they need some kind of AI in their work to keep pace with competitors, but far fewer people understand that input itself makes or breaks a model. Decision-makers, especially in investing, need to ask themselves whether the underlying data is good enough to make a confident call in the first place.
That means questioning the inputs: how a founder slices their data in a slide deck, or filling in the gaps in missing financial information, instead of taking what you have and calling it a day. If your risk assessment still relies on a handful of financial statements and a good pitch, AI does not fix that. It just lets a business that has not yet built resilience tell a more polished story, faster.
What do you think people still misunderstand about your part of the industry?
Investors have a gut instinct that they have honed over decades of successfully distinguishing the future unicorns from everyone else. There are factors that data cannot read, in particular, the qualities of a founding team willing to do anything to make their business a success. As a result, there’s sometimes resistance to AI adoption in certain parts of investment because of the misconception that giving it more control means taking human judgment out of the picture. It doesn’t and it shouldn’t. Good data should augment and inform investors’ judgements, but currently, it is often overridden by instinct.
What do you expect to rise up the agenda over the next year?
Automation and AI will likely change attitudes to risk. Nearly every early growth-stage tech company now builds on some kind of AI, and investors are inundated with AI companies to assess, struggling to separate the wheat from the chaff. There are huge sums flying around, but not every company is being funded. People assume the businesses banks and VCs turn down are ‘too risky’ for institutional funding. In truth, they are often turned down because their full operational picture is invisible to investors. They aren’t equipped to assess their financial value, meaning they cannot assess the risks associated with the capital outlay.
As AI becomes embedded in physical infrastructure, from construction sites to industrial recycling plants, these businesses can look increasingly illegible to an outside observer, even though they are generating more measurable data than ever. Prioritising a real understanding of a company is starting to matter more than it ever has, and I expect more growth capital products to get built specifically for businesses like those.
What’s the one question about your company we should have asked, and what’s your answer?
“Why does the problem your business is targeting need to be solved?”
The answer: Ultimately, good businesses aren’t being funded. Accessing better data is not just about helping investors assess companies more clearly to then turn them down. It’s about everyone being able to see the full picture. When companies have more financial clarity, investors get a clearer idea of the risk involved, which then allows them to deploy more capital to founders who deserve it. It’s a virtuous cycle.
