Intelligent oversight: How LGSCO Is reimagining complaints handling through AI

The Local Government and Social Care Ombudsman (LGSCO) is the final stage for unresolved complaints about local government and social care services in England. As an independent body, their work provides a critical check on public services – particularly for the most vulnerable. Yet behind this vital service, they recognise the need to maintain standards while finding new ways to handle a rising caseload efficiently.

What began as a short consultancy engagement evolved into a trusted partnership. Together, we combined strategy, user-centred design, and AI to deliver a proof of concept that demonstrates two practical tools - ultimately supporting LGSCO in how they might transform their core operational service through the use of intelligent automation.

LGSCO recognises that their case handling system contains some inefficiencies and could benefit from some modernisation. Investigators are often tied up with admin-heavy tasks - like locating documents, redacting sensitive information, and building timelines - time which could be better spent resolving complaints.

The data makes a strong case for change: around 80% of fully investigated complaints are upheld, and over 99% of recommendations to councils are followed. Strengthening early-stage case handling has the potential to deliver faster, fairer outcomes - and gives the team a compelling reason to act.

Although they have limited digital capability and no in-house developers, LGSCO’s leadership leans in. They see these challenges not as a barrier, but as further motivation to rethink their approach. Encouraged by reports like those from the Tony Blair Institute, LGSCO has been exploring how AI can help improve their services - and that’s where we step in.

To date, we’ve partnered with LGSCO through several engagements - beginning with a 10-day consultancy tailored to LGSCO’s starting point. Many participants were new to AI, so workshop sessions were jargon-free, hands-on, and grounded in real case files. Rather than present hypothetical solutions, we worked collaboratively to surface operational pain points and explore where technology might help.

To translate these insights into a usable roadmap, we introduced a framework for prioritising use cases by feasibility and impact. Through live scoring, shared canvases, and collaborative exercises, we built a shared understanding of where transformation could begin. A clear principle was agreed early: AI should support staff, not replace them. By day 10, the team had a common language around AI, prioritised use cases, and received practical literacy guidance along with a roadmap to move forward.

With that foundation in place, we facilitated service design workshops across LGSCO’s core operational areas - intake, assessment, investigation, and information governance. Teams mapped existing workflows in detail, identifying inefficiencies and moments where AI might intervene. These insights formed a practical adoption plan aligned with daily realities and organisational priorities.

The investigation process emerged as the most pressing challenge: investigators spend hours manually searching documents and building timelines, slowing resolution. Based on feasibility and impact, two tools were prioritised - AI semantic search and automated timeline generation - and delivered as a six-week proof of concept in LGSCO’s Azure environment. Investigators began testing them within weeks, demonstrating clear speed to value by easing their biggest daily pain points. We also ran regular workshops and demos to ensure the tools reflected real needs.