Synthetic users panel
Archetype migrations, empirically anchored
For CEE 2030: Energy Taxation, Carbon Pricing, and Households— each segment's population redistributes across archetypes under every scenario, calibrated against country-model demographics and cross-checked by uncertainty bands.
Segments
5Scenarios
4Archetypes
19Profiles
599Emerging customer types
Customer identities that cross your existing CRM segments — no single current product captures them.
Baselines calibrated to Eurostat demographic + EU-SILC; scenarios modeled. Directional signals for scenario planning, not population forecasts.
How your current segments fare across scenarios
For each CRM segment: does it stay intact, drift, or fragment under each scenario?
Simulation Comparison
Markov baseline vs. OASIS LLM-driven multi-agent simulation. Same archetype profiles, same scenario priors, two engines. Divergences ≥5pp are surfaced with interaction-trace classification (signal vs. noise) so consumers can decide whether the LLM layer is adding structural insight or just inference noise.
No simulation artefacts for energetika-a-dane-v-cee-2030-dph-na-elektrinu-co2-zdaneni-a yet.
When the OASIS PoC ships, three artefacts land in this future-space's synthetic-panel/ directory: markov-results.json, oasis-results.json, and comparison.json. The decision criterion is documented in the Week 6 retro journal — GO only if OASIS surfaces ≥2 archetype-migration findings the Markov baseline misses, with traceable interaction logs.