Problem
Taff Housing Association has always put people at the centre of everything it does. But with over 220 colleagues delivering services across South East Wales, the realities of increasing workloads, responding to tenant queries, and keeping up with regulatory changes presented ongoing challenges.
For Taff, administrative burden was a persistent issue. Hours lost searching for the right policy, drafting responses to tenant emails, and keeping up with compliance requirements.
Solution
In 2024, Taff took a bold step: embracing AI to make colleagues’ lives easier and improve outcomes for tenants.
Leading AI’s KnowledgeFlow’s platform was deployed to tackle these pain points head-on. Built securely inside Taff’s Microsoft Azure environment, ensures only Taff staff have access and that sensitive data never leaves the organisation. Its core strength lies in semantic search: staff can ask questions in plain English and the system scans all relevant policies, procedures, and guidance to deliver instant, accurate answers, that can also be translated instantly and perfectly into Welsh.
Examples from the Frontline:
1. Rapid Policy Insight and Contradiction Checking
One of the most powerful features has been the ability to check for contradictions across all policies. As described in a demonstration, “I can check all 108 policies simultaneously… and the AI will flag any direct contradictions, giving a summary of what it’s found.” This has made policy reviews faster and more reliable, freeing up the governance team to focus on improvements rather than endless manual cross-checking.
2. Drafting Tenant Communications
Staff regularly face repetitive queries, such as boiler breakdowns or repairs. Now, when a tenant emails about a faulty boiler, the AI instantly drafts personalised responses, referencing the relevant policies and procedures. “Here’s ten draft responses… you can read through them, add more context, and send them off,” explains one manager. This has saved hours each week and improved the consistency of communications.
3. Tone of Voice and Collaboration
Taff has embedded its own templates and tone of voice guidance into KnowledgeFlow, ensuring that AI-generated documents sound like Taff. Staff review drafts collaboratively with engaged customers, asking, “How does this feel? What’s this sound like? Where do we need to improve?” This blend of technology and human oversight means the AI supports, rather than replaces, professional judgement.
4. Empowering Staff and Reducing Fear
Initial reactions to AI were mixed—some feared it would replace jobs. But as colleagues saw the practical benefits, attitudes shifted. One staff member reflected, “It’s about empowering colleagues, increasing that human connection… freeing up people to do the important work.” The cultural shift has been supported by leadership openly admitting they are “not experts” and encouraging everyone to learn together.
Outcome
Hours Saved: Staff report significant time savings on admin and policy queries.
- Improved Consistency: Responses to tenants and practice across teams are now based on up-to-date, centralised guidance.
- Enhanced Collaboration: AI supports, but does not replace, professional judgement—staff remain in control.
- A More Human Service: With less time lost to admin, colleagues spend more time supporting tenants directly.
Taff’s journey with AI is ongoing, but the results are already clear: less drudgery, more impact, and a team that feels empowered rather than threatened by technology. As one leader put it, “It’s about leveraging the speed that AI gives us, but always keeping people at the centre.”
Taff’s experience shows that when AI is deployed thoughtfully, it can transform not just processes, but the everyday lives of those delivering vital housing services.