OSCAR: Building an AI front door with, not just for, local government
- 1 day ago
- 4 min read
Most councils now accept that “more of the same, but digital” will not close the gap between demand and resources. Yet many leaders are understandably wary of big‑bang technology answers that arrive fully formed from outside and never quite fit the grain of local services.
OSCAR takes a different route. It is an AI‑enabled “front door” for councils being built in partnership with Peterborough City Council, shaped by the realities of their contact pressures, social care pathways and workforce constraints. On the Peterborough side, the work is being led by Julian Patmore, whose team is deliberately treating OSCAR as part of a wider shift in how the council listens and responds to residents, not as a bolt‑on gadget.
Why Peterborough wanted something different
Peterborough had already seen the benefits of domain‑specific AI through earlier collaborations with Datnexa, including tools that support social care and health professionals and a specialist social prescribing assistant. Those deployments showed that AI can behave like a useful junior colleague, structuring information, drafting and signposting, without crossing the line into unchecked decision‑making.
But they also exposed a bigger question: what happens if you apply that “junior colleague” model at the very front door, before residents are sliced into service silos?
Julian and colleagues were clear about the problem they wanted to solve:
Residents rarely describe their lives in neat service categories.
Staff at the point of contact are expected to make complex judgements very quickly, with limited tools.
Each new digital initiative risked creating one more channel with its own rules, rather than a more joined‑up journey.
OSCAR is the joint attempt to answer that problem in a way that feels usable for contact‑centre staff and credible for senior leaders.
What OSCAR actually is
OSCAR is not a single chatbot trying to “know everything”. It is an orchestration layer that sits across access channels and coordinates a set of specialist agents behind the scenes.
From a resident’s point of view, there is one conversational entry point where they can explain what is going on in their own words, whether that is about an older parent who is starting to struggle at home, a family under financial pressure, or someone whose mental health is affecting their housing situation. OSCAR takes that story and quietly does three things:
Breaks it into related needs – for example, potential care and support, community connection, housing stability, money worries.
Routes those elements to the relevant agents – adult social care, social prescribing, housing, children’s services and so on.
Reassembles the responses into a single, coherent reply that makes sense to a non‑specialist.
For staff, OSCAR shows up more like a diligent junior colleague than a new system to “learn”. It can capture structured information from the conversation, draft a summary, highlight possible next steps and surface local offers – all ready for a worker to review, amend and own.
The Peterborough lens: leadership and learning
One of the distinctive features of OSCAR’s development is the way Peterborough is treating it as a live learning project rather than a one‑off procurement. Under Julian Patmore’s leadership, the council is:
Starting where the pressure is most acute, rather than trying to redesign everything at once.
Involving frontline staff early so the assistant reflects real conversations, not idealised process maps.
Testing how OSCAR affects both resident experience and staff workload, with an eye on tangible impacts that finance and performance teams can recognise.
This approach changes the feel of the work. Instead of “installing” a solution, the council and Datnexa are iterating together on where OSCAR sits in the flow of contact, what it is allowed to do, and how its outputs are used in day‑to‑day practice.
For senior leaders, that partnership model matters as much as the technology. It means the tool is being shaped in an environment where safeguarding, budget pressures and political accountability are real, not theoretical.
Practical value: what leaders can expect
While OSCAR is still in development, the direction of travel is clear, and it reflects questions that chief executives, directors and Section 151 officers consistently ask.
OSCAR is being designed to:
Help more people at first contact by dealing with the “whole story” instead of just the presenting issue.
Give staff a better starting point for decisions by automatically structuring and summarising information.
Make earlier, lighter‑touch support easier to access, especially via social prescribing and community‑based options, to reduce the number of cases that escalate.
Provide a clearer line of sight from AI investment to time saved and costs avoided, using the same sort of evidence that Datnexa and Peterborough have generated from earlier tools.
None of this relies on OSCAR making its own high‑stakes decisions. The intent is to free up
human capacity and improve the quality of human decisions, not bypass them.
Built with governance from day one
Because OSCAR is being developed in a live council setting, governance is not a slide that appears at the end of a project; it is baked into the design conversations from the start.
Together, Peterborough and Datnexa are defining and enforcing some hard boundaries:
OSCAR will not diagnose, determine eligibility or make safeguarding judgements on its own.
It will draw on approved local policies, procedures and information sources, with a clear route for keeping those up to date.
It will leave an auditable trail of what was asked, what was suggested and what staff ultimately did.
For senior leaders, this matters for two reasons. First, it keeps ultimate accountability in human hands. Second, it makes it possible to talk about AI in the same way they talk about any other significant operational change: in terms of risk, control and measurable benefit.
Where this goes next
OSCAR is not being built as a Peterborough‑only curiosity. The intention from both sides is to create something that can travel: an AI front door pattern that other councils can adapt to their own structures, policies and partnerships without starting from scratch.
That is why collaboration is so important. By working closely with Julian Patmore and his colleagues, and by grounding design decisions in real cases, real staff feedback and real budget pressures, Datnexa is aiming to produce more than another tool. The goal is a new way for councils to meet residents at the front door: one conversation, one coherent response, backed by technology that understands its place in a high‑stakes public service.
