On Monday, Fifth Dimension announced its Series A, led by HV Capital with continued backing from Seedcamp, Speedinvest and other notable institutional investors. We've invested in the company three times since their pre-seed so this felt like the right moment to set out why we backed them and what we think their journey tells us about the vertical AI opportunity more broadly.

When we wrote about the case for vertical AI in February, Fifth Dimension was a key part of our founding conviction in that thesis. This post is more personal, it’s the story of how that conviction formed, how it was tested and why we think the market is seemingly moving in the direction we and the Fifth Dimension team anticipated.

When Bill and I first met Kate Jarvis and Johnny Morris in late 2022, they weren't pitching an AI company, they were pitching a solution to a problem that Johnny had spent seventeen years living. This was a problem where real estate professionals were drowning in unstructured data, making billion-pound decisions from PDFs and spreadsheets, without technology that spoke their industry’s language.

The combination of Kate and Johnny’s backgrounds (and complementary skillsets) drew us in, as it’s a unique fusion of deep domain knowledge and technical capability. Kate's background in computational linguistics, a Stanford PhD followed by CPTO roles across multiple verticals, gave her the technical depth to see what LLMs could do for domain-specific workflows before the market caught up. Meanwhile Johnny's career running analytics and research at Countrywide, then as CDO at Wayhome, meant he knew precisely which workflows were broken and which buyers would pay to fix them.

At our investment committee in 2023, we noted that the ‘why now’ was particularly compelling. AI was front-of-mind for every exec committee and AI budgets were being forged across every major property firm; the founders who understood the industry's specific data structures would have an insurmountable head start over generic tools trying to adapt.

Our original thesis was that vertically-aligned AI co-pilots would outperform horizontal tools or off-the-shelf models. In fact, even more so in industries with fragmented, proprietary data and complex domain-specific reasoning and real estate was the perfect proving ground.

That thesis has held up well and Fifth Dimension grew their revenue >6x year-on-year in 2025, signing enterprise contracts with some of the largest asset managers in the US and Asia. Usage continues to double quarter on quarter. Multiple team members have been closing significant enterprise contracts independently, proving the sales motion was repeatable and no longer founder-led.

We ended up underestimating the speed at which the market would bifurcate. In our note, we flagged the risk that a sufficiently capable horizontal tool - a Microsoft Copilot or an enterprise ChatGPT wrapper - might close the gap. That hasn't happened; the gap has widened as the product is delivering something that generic AI tools cannot replicate.

We also underestimated how quickly the international opportunity would present itself. Fifth Dimension's revenue is now heavily weighted towards the US, with meaningful traction in Asia. The company we backed as an initial UK play has become a global vertical AI platform. That geographical shift is telling in itself - US adoption of AI in real estate is markedly ahead of Europe, and European operators who wait risk being left behind entirely.

The truth is that we put too much emphasis on appraising the product in its initial use cases e.g. marketing copy work flows for real estate firms. They were still in client discovery mode, though it would have been pretty easy to look at this and underestimate the scale of the opportunity at play - we’re glad we didn’t.

Last week, Fifth Wall's newsletter described the $1.5Bn Anthropic-Blackstone joint venture as a "five-alarm fire" for real estate operators, arguing that companies can no longer rely on external technology providers and must build AI capabilities in-house. Brendan Wallace called it an "HMS Dreadnought moment" - the idea that prior investment in competing approaches becomes a liability overnight.

It's a provocative framing and we think it captures something real about the urgency, however, we'd push back on the conclusion. The Anthropic-Blackstone model - embedding forward-deployed engineers inside operating companies - requires scale, budget and technical infrastructure that most real estate firms do not have. For the vast majority of asset managers, investment firms and property companies, building bespoke AI systems from scratch is neither practical nor economically rational.

This is precisely where vertical AI platforms like Fifth Dimension earn their position. They offer the domain specificity and proprietary data integration that generic tools lack, without requiring each customer to hire a vast AI engineering team. Their moat compounds as every workflow embedded, data source connected and decision logged makes the platform harder to replace and more valuable to the next customer.

There is a strategic question about whether the largest operators - Blackstone, Brookfield, the top ten - will build rather than buy. Some will, however, the long tail of the market where the vast majority of commercial real estate AUM sits, will need platforms built by people who understand their workflows. That's Fifth Dimension's market, and it's enormous.

When we published our vertical AI piece in February, we argued that the most durable value in enterprise AI would be captured by systems that go deep in a specific domain. 

Fifth Dimension's Series A is, for us, partly a validation of that thesis. The company has moved from automating narrow tasks like lease abstraction to delivering an AI-native operating system for real estate decision-making. Their toolkits - pre-built, production-ready AI workflows tailored to specific use cases - are exactly the kind of compound moat we described.

We're proud to have backed Kate and Johnny from the very beginning, and we're looking forward to what comes next.

Adrian Love is General Partner at Love Ventures, an early-stage VC fund investing at pre-seed and seed into Fintech, Future of Work and Consumertech.