
Most conversations about AI in property start with the technology, hunt for a problem to solve, then rush to deploy — even when the data bedrock isn't fit for purpose.
At Ringley, we've always used technology to enrich jobs and support career progression. Our answer to AI is more tech — our own tech: we built our own language learning model, fed it a closed, property-specific brain, so it holds exactly the knowledge we need, without the runaway cost of a third-party AI token.
We're luckier than most, our proprietary tech builds on solid data foundations, not years of back-filling. We tracked where the team lost most time, repetitive processes and tasks, and because our mantra is "bring your personality to work” as our job is to delight customers, any tech that helps alleviate the mundane and empowers employees, is to everyone’s advantage.
Three things shaped our approach:
• We already owned the data — leases, budgets, arrears, compliance, rents, voids and churn across every building we manage.
• Our AI is fully fenced, running in our own environment; no client document or personal data reaches a third-party model provider.
• We'd already automated finance and risk, so we were well on the way to mapping process flows and building predictive systems.
Here's how we use in-house AI for the drudgery and where we still rely on human judgement:
Decisions rely on leases, deeds, consents, planning permissions, S106 agreements, EPC ratings, fire risk appraisals and maintenance certificates. No two leases agree, so the same clause gets re-read constantly, and abstracting a new instruction takes weeks.
We've embedded lease extraction for human acceptance — apportionments, repairing obligations, sub-letting, reserve funds — enabling lease-difference analysis across a portfolio, plus parking schedules and title numbers cross-referenced to HM Land Registry, council tax and EPC records. Once accepted, AI writes it into our database as structured, query able data, each field linked to source — proof AI needs a data rock, not sand. Property managers get answers in seconds, clause attached; solicitors become advisors instead of retrieval engines.
Regular reporting is assembly, not analysis — export, pivot, paste, write the commentary last, describing problems that started six weeks ago. Ringley AI queries our data layer directly, drafts the narrative, benchmarks each scheme against the portfolio, and flags anomalies before anyone asks, always reviewed and signed off by an appointed employee. Analysts get interpretation back instead of pivot tables; managers intervene on arrears and voids. Rigour comes from the system, judgement from the human.
Block management is a precedent industry, held in individual expertise. New joiners can take months to become proficient, while experienced staff field the same questions repeatedly, and there is a risk that expertise leaves when people do.
Our chatbot is trained on two decades of practice, our policies and precedents, plus structured lease and building risk data. It cites sources, distinguishes general guidance from what a lease says, and routes anything weighty to an employee. New starters get a sourced answer at 9pm; senior staff get time back for deep work. Nobody feels judged by a chatbot, so people check rather than guess, which alone prevents errors.
The information needed to find new opportunities is public but scattered and across silos. We use AI to link contacts to companies and projects, turning scattered data into a single view - who to approach, about which building, for which service. AI then works alongside the salesperson, tracking progress, formatting emails, summarising conversations. Because we hold the data, our portfolio insight is marketing nobody else can write.
Anyone can buy an AI model; few in this sector have twenty years of integrated property data. Every use case targets repetitive, low-discretion work, none targets what our people are good at. The AI drafts, extracts and flags; people decide, take responsibility, and make the judgement.
The question isn't "how do we adopt AI?" It's "where are our best people spending unnecessary time?" AI is about elevating efficiency and empowering employees, not replacing them.
Chengpeng Zhao, Data Scientist & Business Analyst, Ringley Group
© 2026 News On The Block. All rights reserved.
News on the Block is a trading name of Premier Property Media Ltd.