Case Study · Live Business Group
Live Insights & Live Stage
Live Business designs and runs entertainment for cruise lines, ferries, hotels and resorts. Guest feedback arrived as spreadsheets and hunches. They wanted structured insight they could act on sailing by sailing, and a reason for a cruise line to choose them.
What we built
Live Insights: after a show, a warm, voice-enabled AI host interviews guests on an iPad kiosk or on their own phone. It draws the themes and a rating out of every conversation and feeds three things: a partner dashboard, an assistant grounded in that operator's own data, and a recommendations engine that refreshes every night. From a guest finishing an interview to its themes appearing on the operator's dashboard takes about 30 seconds.
Under the hood
A multi-tenant platform on Google Cloud with hard tenant isolation, so each operator's guests and data stay separate. Secrets never leave the server (there are no API keys on any device), and a brand-voice guardrail keeps forbidden words from ever reaching a guest. Spend limits, usage monitoring and health checks keep the running cost predictable.
Live Stage
Live Stage, its sister app, ranks a talent roster against a specific sailing's audience so the right acts are booked for the right ship. It went from a client's idea to a working iPad app in ten days and was demonstrated to a UK cruise line at a tender presentation, seeded with the line's own fleet and running on a single iPad with no Wi-Fi.
Outcomes
Eleven demo tenants live. Working software, not slides, in front of a UK cruise line at a tender presentation. Live Insights is on the App Store for iPad and iPhone, and LuminaryAI builds and runs it as a production service for Live Business: the same platform we would build and run for you.
More work
Production AI in Action
Bid Operations
Three cruise-line tenders, one small team, every deadline met
Response plans, deck architecture, clarification questions, RACI and KPI suites, data-drop analysis inside 48 hours, tracked-changes reviews in the team's own Word files, and a slide-by-slide presentation cut that caught a totals mismatch before submission. AI-assisted, human-authored, disclosed to the client.
Financial Services
Six days that found 83–145 recoverable hours a month
A regulated advice, mortgage and accountancy group on a mainstream adviser CRM, cloud accounting and Microsoft 365. Two days on site interviewing every role; a knowledge agent demonstrated in a day; nine use cases in three tiers, a risk register, an 18–24-month roadmap and an ROI model, plus the finding that the biggest gains sat in tools the firm already owned.
Bring us a problem, not a brief.
Discovery call · 45 minutes · online
Tell us where the hours go. We'll tell you what's possible this quarter, what it would cost to run, and whether we're the right people to build it.
or email mattdray@luminaryai.co.uk · 07771 807 283