When a prospective tenant pings a student housing operator at 11pm asking whether a studio in Manchester is still available for the autumn term, somebody (or something) has to answer fast or the lead goes cold. VerbaFlo, a London-based proptech founded in 2024, is betting that the somebody should be a piece of software, and that the software should speak the tenant's language, pull live availability from the property management system, and book the viewing without a human in the loop.
The company says its platform now powers communications and operations across more than 200,000 units globally, and is adding close to 30,000 units a month [Business Review Live, ITBrief, 2026]. In November, VerbaFlo closed a $7 million seed round led by Pi Labs, with participation from Old College Capital, Haatch, and Navigate Ventures [Crunchbase].
The bet
VerbaFlo sells an AI communications layer aimed squarely at residential real estate operators: purpose-built student accommodation (PBSA), build-to-rent (BTR), and multifamily. The wedge is the high-frequency, low-margin work that fills a leasing agent's day: qualifying inbound leads, answering FAQs about pricing and lease terms, sending payment reminders, coordinating maintenance updates, and nudging renewals. The product, marketed under names including VerbaMessenger for web chat and a voice surface for inbound calls, is designed to plug into the operator's existing stack rather than replace it. The company says it supports both native and API-based integration with most leading PMS platforms [VerbaFlo].
The customer list skews toward UK student housing and BTR brands that recognize the operational pain. VerbaFlo cites Homes for Students, Moda Living, Fusion Students, Housing Hand, and Downing among more than 40 client brands it has worked with to refine the platform [The AI Insider, 2026; Free Bible Study Hub, 2026]. A separately announced partnership with Vita Student is bringing voice AI into student accommodation call flows [Proptech Connect].
Why it could be big
The shape of the rental market favors a tool like this. Student housing and build-to-rent are consolidating into the hands of larger operators with thousands of units across multiple cities and, increasingly, multiple countries. Those operators field inquiries in dozens of languages from international students and mobile professionals; VerbaFlo says its platform supports engagement in more than 180 languages [UKTN, 2026]. The unit economics of a human leasing team scale linearly with inquiry volume. The unit economics of an AI agent that handles 80 percent of routine messages do not.
Pi Labs, the lead investor, is one of Europe's most active proptech specialists, and the firm's involvement is a signal that VerbaFlo's thesis (operations-layer AI for residential real estate) lines up with where institutional capital expects the category to land. The North American comparable, EliseAI, has raised at a reported valuation north of a billion dollars selling a similar value proposition to US multifamily operators. VerbaFlo is targeting an analogous gap in the European PBSA and BTR markets, where no clear incumbent has emerged.
Units powered (disclosed) | 200000 | units
Net units added per month | 30000 | units
Languages supported | 180 | units
The team and traction
VerbaFlo was founded by Sayantan Biswas, who serves as Founder and CEO and holds an MS in Data Science from EPFL [LinkedIn, 2026], alongside co-founders Abhishek Garg, VP Singh (COO), and Dan Smith (Chief Business Officer) [LinkedIn, 2026]. The team is distributed across the UK, Europe, India, and the US, a footprint that maps to where its current and prospective customers operate. The seed round will fund expansion of that team and continued investment in the voice and messaging stack.
The traction numbers are the most concrete signal of product-market fit. Going from zero to 200,000 units under management inside roughly two years of company life, while adding nearly 30,000 net new units a month [Business Review Live, 2026], suggests the sales motion into UK and European operators is working. The 40-plus brand customer list includes operators that themselves manage tens of thousands of beds, so each logo carries weight.
What the bears say
The most credible competitive pressure comes from EliseAI, which has a multi-year head start in the US multifamily market and the capital to expand into Europe. Funnel, Knock, LeaseHawk, and Colleen AI also occupy adjacent slices of the leasing-and-resident-engagement stack. Bears will argue that AI leasing assistants are converging on a similar feature set, and that distribution and integration depth, not model quality, will decide who wins each region. The bull answer from VerbaFlo's evidence so far is that European PBSA and BTR are structurally different from US multifamily (shorter lease cycles, international tenant base, different PMS landscape), and that a platform built natively for those workflows, with 180-language support and named PBSA reference customers like Vita Student, has a real local moat. The 30,000-units-a-month run rate is the data point investors will keep watching to judge whether that moat is holding.
What to watch
The next twelve months will turn on three things. First, whether VerbaFlo can convert its UK PBSA stronghold into multifamily and BTR wins on the continent and in North America, where the unit counts are larger and the competition is sharper. Second, whether the WorkFlo product line teased on the company blog evolves into a workflow automation surface that goes beyond messaging into back-office tasks like renewals processing and arrears management. Third, the timing and size of a Series A; a $7 million seed at this growth rate typically buys 18 to 24 months of runway, and Pi Labs-backed proptech companies that hit this kind of unit-growth curve usually return to market sooner rather than later.
Technical breakdown
The architecture VerbaFlo describes is a conversational AI layer (web chat via VerbaMessenger, plus voice) that sits on top of the operator's PMS through native or API integrations [VerbaFlo]. The hard engineering problems are not the language model itself, which is increasingly commodity. They are: real-time PMS state synchronization (availability, pricing, lease status) so the agent never quotes stale inventory; multilingual intent classification across 180 languages where training data is uneven; and graceful handoff to human agents on edge cases without losing conversation context. Voice adds latency budgets measured in hundreds of milliseconds and a much higher cost-per-interaction than text.
What could go wrong at scale
The failure modes worth flagging are operational, not technological. PMS integrations are notoriously brittle; a single schema change at a large PMS vendor can break inventory accuracy across hundreds of properties overnight, and the AI agent will confidently quote wrong prices until somebody notices. Multilingual quality degrades quickly outside the top 20 or so languages, and a mistranslated lease term in a regulated market is a legal exposure, not just a CX problem. Voice infrastructure costs scale with minutes, not seats, so a viral inbound call surge can compress gross margin fast. And the competitive pressure from a well-funded EliseAI moving into Europe is the single largest exogenous risk on the board. None of these are disqualifying. All of them are the kind of thing that determines whether 200,000 units becomes two million, or stalls.