--- title: "Automating a ~600-person care provider, one governed workstream at a time" description: "A multi-branch supported-living provider's whole digital spine, rebuilt as ~18 governed workstreams — with a live ERP that turned messy CSVs into 294 clean carer records and 588 auto-created supervision schedules." url: "https://automancer.uk/work/care-provider-transformation/" canonical: "https://automancer.uk/work/care-provider-transformation/" type: "CaseStudy" datePublished: "2026-06-20T00:00:00.000Z" --- # Care without the clipboard. We're rebuilding the entire digital spine of a multi-branch, CQC-regulated care provider — around 100 office staff and 500 support workers — as roughly eighteen governed workstreams, several already live in production. - **294 + 293** — carers and service users imported as clean, canonical records from real CSVs - **588** — supervision & spot-check schedules created automatically - **274 / 418** — duplicate/ambiguous records auto-resolved by identity matching - **0 / 0** — failures and warnings across the production infrastructure audit ## A whole care business, held together by spreadsheets and goodwill Supported-living care runs on paperwork — rosters, supervisions, medication records, training compliance, incident logs, mileage claims, timesheets — and this provider was running most of it by hand, across three branches, on a patchwork of outsourced platforms and spreadsheets that didn't talk to each other. The costs of that show up everywhere a regulator looks. Compliance evidence scattered across different providers. Refresher training tracked manually. Care-quality signals buried in separate systems that only got cross-checked when someone had a spare afternoon. And a 500-strong support-worker workforce that didn't even have a consistent digital identity to log in with. They didn't need one more app. They needed someone to systematically automate the whole business — carefully, in the right order, without breaking the care in the meantime. ## One transformation, delivered as ~18 governed workstreams Instead of a risky big-bang overhaul, we broke the transformation into governed lanes — each scoped, built, and board-gated before anything touched real client data. Here are the ones that matter most. ### The core platform (ERP / CRM) The strongest single piece of the work. We built a modular platform to replace the reach of an outsourced system that only covered recruitment — a canonical people registry with identity matching, an append-only audit trail, an ingestion pipeline (email, automated capture, manual entry), report mirrors reconciled against real exports, a task engine with a "My Planner" view, service-user referral-to-onboarding journeys, incident case management, and a requisition-governance module. The hard facts, all from the real work: - A single roster import created **294 carers and 293 service users** as clean canonical records — straight from the client's real CSVs. - Identity matching **auto-resolved 274 of 418** review candidates, so humans only had to look at the genuinely ambiguous ones. - **588 supervision and spot-check schedules** were instantiated automatically. - Everything reconciled against real exports — 500 visits, 175 medication alerts, 416 names routed to review — so the numbers could be trusted. It's live: a production pilot the client is actively using and giving structured feedback on. ### Self-hosted Learning Management System We replaced outsourced training and manual, spreadsheet-based refresher-tracking with a self-hosted LMS — plus a custom compliance layer to track mandatory, refresher and overdue training the off-the-shelf tool couldn't handle natively. Seeded across all three branches, courses loaded, and staged for go-live. It turns scattered CQC inspection evidence into one place you can actually point an inspector at. ### QA / management oversight dashboard Live. It pulls care-quality signals from multiple systems into one view — KPI cards, branch comparisons, a support-worker priority watchlist, a repeat-concern watchlist for service users, medication review, and a filterable review queue. Instead of hunting through separate systems for exceptions, managers see what needs attention in one place. ### Production infrastructure, done properly All of this runs on hardened, monitored, EU-hosted infrastructure built as code — with real backups and a *verified* restore test, not just a hopeful one. The documented audit came back with **zero failures and zero warnings across every layer.** We don't just build it; we run it properly. ### Support-worker digital identity The 500-strong support-worker workforce had no consistent login identity — which was quietly blocking the LMS and portal rollout. We gave every worker a managed, EEA-hosted mailbox identity, with DNS verified and provisioning tooling built to roll it out safely. **And more, in active build:** a staff intranet and portal (announcements, HR resources, policy acknowledgement, clock-in/out), a mobile mileage-claims app with an automated eligibility-and-variance rules engine, and support-worker leave requests — plus a stack of discovery work: an M365 governance audit, an AI rota-scheduler feasibility study, VoIP call-transcription compliance analysis, and governed WhatsApp announcements. ## Governed, staged, and never reckless with real care data Care is not a place for "move fast and break things." Every workstream was governed: scoped, built on staging, reconciled against the client's real exports, and board-gated before it went anywhere near live service-user data. The client tests each lane and gives structured feedback before it's signed off. That's the same discipline we hold our own infrastructure to — audited, monitored, backed up, restore-tested. It's why a care provider can trust us with the spine of their business. ## Live, in production, and growing - The core platform is a **live production pilot** the client is actively using. - **294 carers and 293 service users** exist as clean, canonical, auditable records where before there were CSVs and duplicates. - **588 supervision schedules** are created and tracked automatically instead of by hand. - The QA oversight dashboard is **live**, giving managers one place to see what needs attention. - Production infrastructure passed its audit with **zero failures and zero warnings.**