Simulate
10,000 people.
Before asking one.
VDC builds synthetic social populations that think, react, and decide — so you can test policies, messages, and products on a digital society before touching the real one. No personal data. No surveys. Results in hours.
(Stanford, 2024)
first simulation results
collected or stored
synthetic by design
Decisions that affect millions — tested on dozens
Every policy rollout, product launch, and communication campaign is a blind bet. Traditional research is slow, expensive, and limited by sample size. By the time you get results, the window is closed.
- Focus groups: 8–12 people, 3–6 weeks, $30k+
- Surveys require personal data consent — GDPR friction
- Can't simulate cascading social reactions over time
- No way to test second-order effects before launching
A synthetic society you can stress-test anything against
VDC generates calibrated digital populations — AI personas grounded in real demographic, psychographic, and behavioural patterns — that you interact with like real focus groups, at scale, instantly.
- Simulate 1,000–100,000 personas in minutes
- Zero personal data — GDPR compliant by architecture
- Track how reactions evolve across time and events
- Test 10 scenarios in parallel; see which one wins
From brief to insight in 48 hours
Three steps. No data collection. No waiting for fieldwork. No statistical noise from tiny samples.
Design your population
Specify the demographic, geographic, and psychographic profile of the society you want to simulate. Our engine generates a calibrated synthetic population — grounded in public data patterns.
Ask them anything
Run queries, expose personas to policy scenarios, messages, or events. Each synthetic person responds based on their profile, history, and social context — not a random sample.
Read the society's reaction
Get segmented breakdowns, sentiment maps, and predictive trajectories. Understand not just what people think — but why, and how opinions will shift as conditions change.
Why VDC, not the others
| Capability | VDC | Simile.ai | Unanimous AI | Metaculus |
|---|---|---|---|---|
| No personal data required | ✓ Synthetic-native | Requires real interviews | Requires live users | Requires live users |
| GDPR compliance by default | ✓ Architecture-level | Post-hoc anonymisation | Complex consent flows | Partial |
| Scale to 10k+ personas | ✓ Minutes | Limited by real panels | Swarm size-limited | Forecasting only |
| Time-evolving social dynamics | ✓ Multi-round sims | Snapshot only | Real-time, not predictive | Aggregated trends |
| Deploy in under 48 hours | ✓ Self-serve + API | 6–8 week enterprise onboarding | Weeks | Days |
| SMB / mid-market pricing | ✓ Accessible tiers | Enterprise-only ($100k+) | Enterprise-only | Freemium |
| Policy / govtech use cases | ✓ Core vertical | Secondary | Not positioned | Adjacent |
Who runs simulations on VDC
Anywhere you need to understand how a population reacts before you act.
Government & policy
Test public reactions to housing reform, tax changes, or health mandates on synthetic citizens — before the parliament vote.
Communications & PR
A/B test messaging at scale. Predict how a press release lands across demographics before it reaches journalists.
Product research
Simulate market adoption curves. Understand adoption barriers and brand perception without a survey panel.
Academia & think tanks
Run large-scale social experiments ethically. Replicate or extend studies like Stanford Smallville at a fraction of the cost.
Pharma & public health
Model vaccine uptake, medication adherence, or epidemic responses across synthetic population segments.
Urban planning
Simulate resident reactions to infrastructure projects, zoning changes, or mobility policies — before the budget is approved.
This works. The research proves it.
VDC's architecture is grounded in published experiments validating digital social simulation.
1,052 real people — 85% replication accuracy
Generative AI agents, when calibrated with demographic profiles, replicate survey results from the General Social Survey with 85% accuracy across opinion questions.
25 AI agents form an emergent social world
Agents with memory, reflection, and planning spontaneously organised social events, spread information, and formed opinions — without scripted behaviour.
Digital twins for dropout prevention in education
Estonia's Social Digital Twinner project used AI population models to identify at-risk students and simulate policy interventions — before real implementation.
Simile builds clones of real people.
We simulate entire societies.
Problem
Every significant decision — from government policy to product launches — is tested on tiny, slow, expensive samples of real people. The research industry is a $90B market that hasn't changed in 40 years.
Solution
VDC generates calibrated synthetic populations — AI-native social twins not based on any real individual — that organisations can query like a living focus group. 10,000 personas in 10 minutes. Zero personal data. Full GDPR compliance by architecture.
Differentiation from Simile.ai
Simile clones real people (requires interviews, consent, real data pipelines) and serves Fortune 500 enterprise at $100k+ contracts. VDC is synthetic-first — no personal data required — making us accessible to governments, mid-market companies, and researchers globally. Where Simile needs 6 weeks onboarding, VDC delivers in 48 hours.
Market entry
Govtech and policy research is our beachhead: high willingness-to-pay, urgent need, and regulatory alignment (governments can't use systems that process citizen data). From there: pharma, communications, product research.
Why now
Stanford's 2024 study validated synthetic agent accuracy at 85%. The EU AI Act (Aug 2026) will make real-data approaches more expensive and complex. The regulatory window is closing for our competitors — and opening for us.
Run your first simulation.
No data needed.
Join 40+ organisations on the early access waitlist. First pilots launch Q3 2026.
No spam. No commitments. We'll reach out personally.