How Tomas stopped paying humans to answer the same 10 questions.
Key results — VALIRIS deployment, UK portfolio (2025–2026)
Before VALIRIS, Tomas had a system. It wasn't pretty, but it worked. Sort of.
Managing over 50 short-term rental properties in the UK, he'd built a small team of virtual assistants based overseas to answer guest messages. Check-in questions. WiFi codes. Late check-out requests. The same ten questions, 50 different ways, arriving at all hours.
It cost him roughly £10,000 a year. And it still wasn't quite right.
The VAs were good. But they weren't instant. They weren't answering at 3am. They occasionally got things wrong. And every week, Tomas was still personally fielding the ones that slipped through.
A bespoke automation, hand-built for one client's operation.
We connected everything Tomas was already using: his booking platforms, his property data, his guest messaging channels. The AI ran through n8n workflows. Conversations were tracked in Google Sheets. Notifications fired through Slack. It was stitched together, purpose-built, and not pretty to look at from the inside.
We started on a single property. Then two. Then three. Tomas's team worked through the portfolio one property at a time, building out documentation for each. We fixed issues as they surfaced. We didn't rush it.
By the time we were happy, we were live on all 50.
Since going fully live, VALIRIS has handled over 6,000 guest messages across Tomas's portfolio. More than 90% automatically, with no human involvement. Average response time: under 30 seconds.
The VAs were no longer needed.
In that same period, VALIRIS didn't just save time. It made money. The AI proactively offered extras at the right moment: late check-outs, early check-ins, additional services. In February 2026 alone, just five properties generated £1,400 in net extras revenue from AI-sent payment links. Without Tomas doing a thing.
Watching 6,000 messages flow through a system built from connected spreadsheets, Slack channels, and n8n workflows, one thing was clear: it worked better than the VA team it replaced, at a fraction of the cost, and guests had no idea anything had changed.
And we kept thinking: no STR operator should have to build this from scratch or run their guest communications from a Google Sheet.
That became VALIRIS. Everything that was scattered across tools (the conversations, the knowledge base, the extras, the analytics, the process rules) now lives in one platform, built for the operator, not the developer. No Slack required, no spreadsheets, no stitching.
The part that still makes us smirk.
Times guests asked "is this AI?" Across 6,000 messages.
That's not a number we put in a pitch deck. It's proof that what we built works.
Valiris has been brilliant across my properties – honestly can't believe how well it's performing.
| Results at a glance | |
|---|---|
| Properties live | 50+ |
| Guest messages handled by AI | 6,000+ |
| Automation rate | 90%+ |
| Average response time | Under 30 seconds |
| Extras revenue (Feb 2026, 5 properties) | £1,400+ |
| Annual VA cost eliminated | ~£10,000 |
| Times guests asked "is this AI?" | 2–3 |
Tomas's portfolio is still running. Still automated. The difference now: you don't need a developer, a Slack workspace, or a Google Sheet. You need VALIRIS. If you manage 5 or more properties on Lodgify, you can be set up the same way. Start with one property. Get it right. Then scale.