What if your data could finally answer back?

Explore the answers — without a single patient record leaving your organisation.

Meet Explore.
Now you can have evidence

for every question.

Alex Rivers

Hello Alex!

Use the right side to interact with AI. Below you can see the results.

☷▦
Top 5 Most Common Diseases
patient_count
4,000 3,000 2,000 1,000 0 Hypertension T2 Diabetes COPD Osteoarthritis Ac. bronchitis
Acute bronchitis · patient_count 1,877
AI AssistantSQL
This query groups patients by mapped diagnosis and counts unique patients per condition — so someone diagnosed twice is only counted once. It excludes unmapped conditions, then sorts by patient count and returns the top 5.
Suggestions for next steps:
  • Expand the list — see the top 10, 20 or 50 conditions
  • Filter by time period — narrow down to a specific year
  • Drill into a condition — build a feasibility cohort from it
Show me the 5 most common diagnoses in our patients↑
Alex Rivers

Hello Alex!

Use the right side to interact with AI. Below you can see the results.

☷▦
Age Distribution — Diabetic Cohort
% of cohort
34% 65–74 yrs · largest band
<45 · 18% 45–64 · 29% 65–74 · 34% 75+ · 19%
65–74 yrs is the largest band · 34% of the cohort
AI AssistantSQL
I split your diabetic cohort into four age bands and counted unique patients in each. The 65–74 group is the largest, at just over a third of the cohort — relevant if you're weighing eligibility criteria around age.
Suggestions for next steps:
  • Cross-tab by sex or comorbidity burden
  • Compare against your site's general population
  • Export this breakdown as a feasibility snapshot
What's the age breakdown of our diabetic cohort?↑
Alex Rivers

Hello Alex!

Use the right side to interact with AI. Below you can see the results.

☷▦
GLP-1 Adherence — 12 Month Trend
% adherent
100% 80% 60% M0 M3 M6 M9 M12
Month 12 adherence · 61% (↓31pp vs. baseline)
AI AssistantSQL
Adherence to GLP-1 therapy drops steadily after initiation — from 92% at month 0 to 61% by month 12, a 31 point decline. The steepest drop happens between month 3 and month 9.
Suggestions for next steps:
  • Segment the drop-off by prescriber or site
  • Flag patients below 70% adherence for outreach
  • Compare against a different GLP-1 formulation
How has GLP-1 adherence changed over 12 months?↑

From "what if?" to "now we know".

Study feasibility
How many eligible patients exist — and where are they?
18,650
eligible patients identified across 6 countries
Clinical outcomes
What happens after treatment leaves the trial?
Mo. 3 94%
Mo. 6 87%
Mo. 12 79%
Mo. 24 68%
Provider intelligence
Where are the care gaps inside our own organisation?
18 pathways monitored · gap closing across every site
Rare disease
Can we see a disease journey that no single site can show?
312
vs. fewer than 40 typically visible at any single site

Explore The Evidence Network

•

Explore The Evidence Network •

Coming next

The next chapter of Explore.

We’re developing new ways to plan studies, build cohorts and manage data access.

From protocol to query

Study protocol ingestion

Turn study protocols into structured research queries.

ClinicalTrials.gov connection

Use registered trial information to support trial replication workflows.

Build research cohorts

Cohort builder

Define patient groups against study criteria to support feasibility assessment and future recruitment workflows.

Privacy & governed access

Privacy risk scoring

Support the assessment of privacy risks in research workflows.

Data access management

Link data access to approved study permits through DAAMS, with a traceable record of access.

These capabilities are in development and are not yet available to customers.

For Healthcare Organisations

Ask your own data.

Populations, pathways and outcomes, without waiting on a data team.

  • Pathway variation across your population
  • Adherence and outcomes by cohort
  • Evidence for quality improvement

Ask your first question.

For Life Sciences

Test feasibility first.

Traceable answers from real-world data, before you commit.

  • Eligible cohorts before the protocol
  • Site populations vs. enrolment targets
  • Real-world outcomes vs. trial data

Challenge Explore.

  • Data stays at sourcePatient-level records never leave the organisation.
  • Every answer traceableEach result shows the exact query behind it.
  • Your governance appliesEvery use runs under the partner's own approvals.
  • ISO 9001
  • ISO/IEC 27001
  • ENS
  • Cyber Essentials Plus
  • NHS DSPT

One question.
A whole network answering:
The Evidence Network®

The Evidence Network
Hospitals
OMOP Standard
Registries
Researchers
Health Systems
Life Sciences
Longitudinal Data
Federated Data
90M+ Lives represented
400+ Healthcare organisations
Billions of health data points
10+ years building trust across health systems

Explore runs on The Evidence Network — a federated ecosystem connecting healthcare organisations without moving or centralising their patient data, structured on the OMOP common data standard.

Meet POLLY · Built by Promptly

Different hospitals.
A shared understanding of the data.

POLLY — the Promptly Ontology Layer — is Promptly’s clinical knowledge layer. It is designed to help Explore apply consistent definitions of patients, treatments and outcomes across different data sources.

What counts as a patient on treatment?

Anyone with a recorded prescription, or someone receiving treatment during a specific study period? POLLY is designed to make that distinction explicit, so comparisons use consistent criteria.

Define

What are we measuring?

Designed to make the criteria behind patients, treatments and outcomes explicit.

Reuse

Are we using the same criteria?

Designed to reuse clinical definitions across studies and data sources.

Review

Can we check the definition?

Designed to support clinical review, with definitions linked to their sources and version history.

Patient-level data stays at its source.

Security and privacy have always been central to Promptly. Explore is built on that same commitment.

Federated by design

Queries run within your organisation’s infrastructure. Patient-level records stay in your environment.

Privacy-first, always

Raw clinical records are not passed to the AI model. Your organisation’s governance remains in place.

Answers you can check

Review the query behind each result. AI-assisted outputs require professional review and interpretation.

Promptly’s certifications & assurance

  • ISO 9001
  • ISO/IEC 27001
  • ENS
  • Cyber Essentials Plus
  • NHS Data Security and Protection Toolkit
Learn more about our security and compliance

See Explore in action.

(Turn on the sound.)