The registry is built. The work around it still runs on people and inboxes.
Alvento is a UK AI operations studio. We build and run AI systems for clinical registries, national clinical audits, royal colleges and healthcare professional associations: submission checking, report drafting, member enquiries and document search, with a human signing off everything that leaves the building.
Most of a registry team’s week is not in the database
Registry platforms store the data. General AI agencies sell chatbots. The work in between is usually still done by hand, by a small team, against deadlines set by a commissioner or a regulator.
Submissions arrive incomplete
Missing fields, implausible values and duplicate records are found weeks later, during analysis, when chasing the submitting unit is slowest.
Reports are rebuilt every cycle
Annual reports, outlier notifications and unit-level summaries follow the same structure every year, yet the first draft is still written from scratch.
The inbox is the service desk
Members and contributors ask the same questions about CPD, revalidation, data definitions and deadlines. Staff answer each one by hand from documents that already exist.
Guidance is hard to find, even internally
Curricula, data dictionaries, SOPs and committee papers live across shared drives and PDFs. Finding the current version is a job in itself.
If the output can’t be audited, it doesn’t ship
Clinical registries and professional bodies answer to commissioners, regulators and their own members. We build AI to the same standard as the record itself.
A named human approves every external output
AI drafts, checks and sorts. People at your organisation approve anything that reaches a member, a contributing unit or the public.
Every answer traces back to a source
Answers and drafts cite the document, dataset version or record they came from, so a reviewer can check the claim instead of trusting it.
An append-only log of what the system did
Inputs, outputs and approvals are recorded as they happen. Corrections are new entries, never silent edits.
Patient data stays where your governance says
We design around UK GDPR, your data-sharing agreements and your DSPT position from the start, and keep identifiable data out of any process that does not need it.
Five places AI earns its keep in a registry
Each one is scoped to a single workflow, built inside your environment, and measured against the hours it gives back.
Submission quality checks
Incoming records are checked against your data dictionary as they arrive. Missing fields, out-of-range values and likely duplicates go to a review queue with the reason attached, while the submitting unit can still fix them.
First drafts of audit and annual reports
Unit summaries, outlier letters and report sections are drafted from the approved dataset, using last cycle’s structure and wording, with each figure linked to the query that produced it.
Member and contributor enquiries
Emails about CPD, revalidation, data definitions and submission deadlines are sorted and answered in draft from your published guidance. Staff approve, edit or take over.
Search across your own documents
An internal assistant for curricula, data dictionaries, SOPs and committee papers that answers with a citation to the current version, not a paraphrase from memory.
Accessible, findable public information
WCAG 2.2 and PSBAR accessibility, plus the structure that lets search engines and AI assistants quote your guidance correctly. Start with our free website audit.
Immunaris, a clinical registry built to be inspected
Immunaris is our reference vaccination registry. It shows how we build regulated record systems: compliance fields enforced by the schema, an append-only signed ledger of clinical events, and clinic, cohort and role isolation enforced in the database. It is an Alvento-owned reference artefact, not a client deployment.
Tell us about one workflow
Pick the task that takes your team the most hours each month. Send us a short description and we will reply in writing with how we would approach it, what it would need from your side, and a fixed-scope quote. Enquiries are answered in writing.
FAQ
Do you replace our registry platform?
No. We work alongside the platform you already have. Our systems sit around it, handling the checking, drafting, enquiries and search that currently happen by hand.
Can AI be used on patient-identifiable registry data?
Only where your governance allows it, and often it is not needed. Most submission checks, report drafts and enquiry handling can run on pseudonymised or aggregate data. We design each system around your data-sharing agreements and DSPT position before any build starts.
Who is accountable for what the AI produces?
Your organisation, through a named reviewer. Every external output is approved by a person, and every draft links back to its sources, so the reviewer can check it quickly and the approval is recorded.
Do you work with royal colleges and membership bodies, not only registries?
Yes. Member enquiries, CPD and revalidation questions, curriculum and guidance search, and accessible public information apply to royal colleges, specialty associations and healthcare professional bodies as much as to registries and national clinical audits.
How does an engagement start?
With one workflow. You describe it in writing, we reply with an approach and a fixed-scope quote, and the first system is built and measured against the hours it saves before anything else is added.
Who is Alvento?
Alvento Ltd is a UK company (Companies House 16394298), founded and run by Oscar Espinoza. We build and operate AI systems for regulated organisations and do not outsource client work.