The clinical data platform Verita's AI is built on.
Verita Neuro, part of Verita Healthcare Group, treats people with spinal cord injuries using electrical stimulation. What makes the treatment work is knowing which electrodes to fire for which movement, and Verita's therapists have years of that expertise.
We built Verita a data collection platform that captures every session as clean, accurate, structured data, across every clinic and every therapist. On top of it sits a model that suggests electrode configurations and gets sharper with each record added.
“Intersect AI learned our complex procedures and found non-intrusive ways to capture the data, so we can finally apply AI to deliver better patient outcomes.”
Julian AndrieszCEO, VeritaYears of clinical expertise, ready to put to work.
Every patient goes through pairings and mappings: sessions where a therapist works out which electrode configuration produces which movement, and how that changes over weeks of therapy. Those records hold the clinical know-how of the whole company. Verita wanted to capture them cleanly and the same way every time, so AI could learn from them and every clinic could benefit.
One shape for every session
A single structure for recording sessions across every clinic and therapist, so the data is ready for a model to learn from.
Senior know-how, shared
Put the experience of Verita's senior therapists in front of every therapist setting up a configuration.
Every case, searchable
Let a therapist preparing a session see what worked for similar patients in any clinic.
Ready for more clinics
Grow into new locations with records captured at the point of care and very little admin per clinic.
A data collection platform made for clinicians, ready for AI.
Every field maps to how therapists already work, and every record lands in the same clean structure a model can learn from.
Session records in one structure
Patients, visits, programmes and stimulation parameters, plus an electrode grid therapists click to record a configuration. Same fields, same format, every clinic.
Weekly updates everyone can read
Summary, progress, challenges and plan for every patient every week, so any therapist can pick up a colleague's case.
Older records brought in
An import pipeline reads earlier records into the same structure, so the patient history counts too.
Configuration suggestions
Therapists choose the muscles to target and see suggested electrode configurations and frequency. A tabular transformer model learns from each new record, and therapists log whether a suggestion worked and why.
Privacy designed in
Personal details, health records and metadata classified and stored separately, a storage region chosen per patient, encryption, access logs and audit trails.
Works on a phone
Therapists check a patient's history between sessions on their own device while the data stays in one central place.
Built around how therapists already think.
We sat with the therapists before designing a single field. Recording a session follows the steps they already work through, in the same order.
One record per patient, from first visit to final mapping.
Data a model can learn from
Mapping and pairing data sits in one place, structured the same way across clinics. Every AI project Verita runs from here starts on it.
Records made at the point of care
Therapists record sessions directly in the platform, so each new clinic adds patients with very little extra admin.
Juniors learn from seniors' cases
Therapists review and revise their own patients' progress, and every record doubles as teaching material for the people they mentor.
New clinics start with the knowledge
A clinic opening next year gets the stored experience of Verita's senior therapists from its first patient.
A base for published research
Clean, consistent records give Verita a foundation for clinical studies and peer-reviewed papers.
With the data flowing, the next AI layers are planned.
Two more layers are planned on top of the platform, both fed by the clean data now flowing in from every clinic.
- A therapist describes a problem in their weekly update and gets a list of similar past cases back.
- Plain-English questions over the full history, for therapists, managers and researchers.
- A copilot that helps write the weekly update.
- A prediction panel that ranks candidate configurations before a session.
- Every prediction logged against the configuration the therapist actually used.
- A review dashboard showing where the model is right and where it needs more data.
Sitting on data you can't use yet?
Tell us what you want AI to do and where the data lives today. One of the senior team reads every enquiry and replies with how we would approach it. Intersect AI is a certified member of the Claude Partner Network.
Rather talk it through? Book a call with the team
You'll speak to the senior team from the first call.