Interactive demo · DCBR Pipeline

See how Epistamate scores a claim.

Enter a research question. The pipeline extracts claims, computes evidence-quality confidence scores, runs adversarial challenge, tracks gaps, and produces a structured brief — not a summary.

Suggested questions →
Carbon border adjustment mechanisms
AI regulation effectiveness EU vs US
Evidence base for four-day work week policies
Generative AI impact on knowledge worker productivity
General research
Policy / Think tank
Consulting
Regulatory / Compliance
About this demo: This demonstration simulates the Epistamate DCBR pipeline — showing how claims are extracted, scored against evidence-quality criteria, challenged adversarially, and compiled into a structured brief with gap tracking. In the full desktop application, this runs locally with multi-provider parallel retrieval, a persistent knowledge graph, and an immutable Decision Log. Evidence scores reflect the formula structure; in production, source tier assignments draw on real document ingestion and retrieval.