Oracle Digital Twins

The digital twin dashboard

Interrogate a cohort. Get answers in seconds.

Digital twins are interrogatable, persistent profiles of a population. Define a cohort — physicians, patients, consumers, or a custom private domain built for you — and ask it questions through one dashboard. Each domain runs on its matched Oracle model, so every answer is grounded in the right population.

Live Interrogation — HCP domain · Oncology cohort · n=250
ASK

Among community oncologists, how likely is adoption of a lower-toxicity bispecific with more complex administration?

66%
Likely to adopt
19%
Neutral / need data
15%
Unlikely
Drill deeper → "Of the 19% who need data — what evidence moves them?"
Any domain
HCP, patient, consumer, or custom
Seconds
From question to readout
Persistent
Reusable, coherent cohorts
Per-seat
Subscription, not per study
What you can do

Define a cohort once. Interrogate it from every angle.

Every follow-up reuses the same locked audience, so you can keep drilling without re-fielding anything.

Ask & quantify

Pose any question and get a calibrated probability distribution back — single query, multi-choice, or numeric — with answer options generated for you if you want.

Crosstab & compare

Break results down by one or two segments at once, or run a battery to compare options side by side against the same cohort.

Drill into any slice

Chain conditional follow-ups scoped to any responder slice — "of the 18% who'd defer, why?" Each twin answers coherently with its own prior responses.

Domains

One dashboard. The right model for each population.

The product is the twins; the model is plumbing that follows the domain. Pick the population that matters to you.

● Live
Healthcare (HCP)
Powered by the Oracle HCP model

66,503 physician digital twins built from real CMS data, across six live specialty modules — Cardiology, Oncology, Primary Care, Endocrinology, Rheumatology, and Hospital Medicine. Interrogate adoption, brand, message, and barrier logic.

● Live
Custom / Private
Powered by the matched Oracle model

A twin cohort stood up for a specific customer's population and kept private to them. Bring the audience that matters to your business and interrogate it on demand.

○ In development
Patient
Powered by the Oracle patient model

Persistent patient cohorts for treatment attitudes, journey, and experience. The underlying patient model is live; cohort instantiation is being built out.

○ On request
Consumer & Social
Powered by the consumer & social models

The same dashboard runs on Simsurveys' consumer and social research models. Talk to us about standing up a consumer or public-opinion twin cohort.

How it works

From cohort to readout in four steps.

1

Pick a domain & cohort

Choose the domain, then narrow by segment — specialty, setting, demographics, region. The dashboard draws a representative, inspectable sample of twins.

2

Ask your question

Type it conversationally or paste a questionnaire. Single query, crosstab, or battery comparison — answer options can be generated for you.

3

Read & drill

Get a probability distribution as a chart or table in seconds. Chain follow-ups against the same locked audience to go deeper.

4

Persist & export

Every question becomes a reusable result card inside a saved project. Export to PDF, PowerPoint, Excel, or CSV.

Example interrogation

A real question, answered in seconds.

Pose a question to a defined cohort and the dashboard returns a calibrated distribution — then lets you interrogate the segments behind it. Illustrative readout from an HCP-domain Endocrinology cohort (MD · NP · PA).

  • Probability distributions, not anecdotes
  • Segment indices show who over- and under-indexes
  • Follow-ups stay scoped to the same respondent universe
  • Each twin's record is coherent — answers stay consistent with its prior responses
Customer Decision Logic · HCP · Endocrinology
"Would you start tirzepatide on a Type-2 diabetes patient whose A1c is still above target on metformin?"
85%
Yes — directly or in specific patient profiles
9%
No — would try another agent first
6%
Need more info / would not prescribe
Surfaced in under 5 seconds from a defined Endocrinology cohort.
Why digital twins

Traditional research vs. the twin dashboard.

Traditional Digital Twins
Timeline Weeks per study Seconds per question
Cost model Per study, per field Flat seat license
Recruitment Hard-to-reach, costly incentives Instant, precise targeting
Follow-up New field, new wait Drill the same cohort live
Continuity One-time snapshot Persistent, reusable cohorts
Traceability Black-box panel Grounded, inspectable twins
Validated, not asserted

Accuracy you can check.

The dashboard delivers population- and segment-level intelligence, validated against live studies. It's a research acceleration and triage tool — fast enough to ask a hundred questions, grounded enough to act on the answers.

  • 9 published validation studies across healthcare, consumer, patient, and retail
  • KL divergence in the 0.05–0.10 range across studies
  • No protected health information (PHI) in training data
See the Validation →
Built for trust
9
Published validation studies
0.05–0.10
KL divergence range
4
Domain models
No PHI
In training data

Frequently Asked Questions

What are Simsurveys digital twins?

Digital twins are interrogatable, persistent profiles of a population. You define a cohort and ask it questions through the dashboard, getting back calibrated probability distributions in seconds. Each twin's record stays coherent — its answers remain consistent with its prior responses.

How does the dashboard work?

The dashboard is how you use the twins. You build a cohort by domain and segment, ask questions conversationally — single query, crosstab, battery comparison, or conditional follow-ups — and read the results as charts or tables. Every question becomes a reusable result card inside a saved project, exportable to PDF, PowerPoint, Excel, or CSV.

What domains are available?

Healthcare (HCP) and custom private domains are live today. A patient domain is in development. Each domain runs on its matched Oracle model — the HCP model for physicians, the patient model for patients, the consumer model for consumers — so the intelligence is grounded in the right population.

What is a custom private domain?

A custom domain is a twin cohort built for a specific customer's population and kept private to them. We stand up the domain against the appropriate Oracle model and your defining inputs, so you can interrogate a population that matters to your business — contact sales to scope one.

How is it priced?

The Digital Twin Dashboard is licensed as a seat-based subscription for teams running ongoing intelligence, rather than per-study. Pricing is custom to seats and domain scope — contact sales for a quote.

Interrogate the twins yourself.

Bring a real question. We'll put it to a cohort live and drill into the logic behind the answer. The dashboard is licensed per seat — talk to us about domains and access.