Computing over evidence, disagreement and judgement, by Pivot AI

Pivot AI’s V3.5 architecture is designed for adversarial matters, such as financial disclosure, where conflicting claims, incomplete evidence, professional judgments, and changing documents must remain traceable over time. It models each contribution as a separate, attributed assertion linked to a stable subject, source version, date, and provenance, allowing competing values, such as a client’s declared account balance and a bank statement’s recorded balance, to coexist without prematurely resolving the conflict. Explicit rules generate findings about discrepancies and evidence gaps, while professionals make separately recorded, purpose-specific decisions that preserve the underlying alternatives and rationale. SurrealDB provides a graph-based foundation for linking assertions, sources, dependencies, decisions, and outputs, enabling the system to identify exactly which work is affected when new evidence arrives. Shared domain handlers govern actions across client portals, professional workbenches, agents, and background workers through typed, versioned operations, scoped permissions, approval controls, and audit receipts. The approach also distinguishes persistent matter state from agent memory, workflow execution, and readable summaries, while using durable recovery and idempotency mechanisms for uncertain external operations. Overall, the architecture aims to maintain continuity, authority boundaries, provenance, and reviewability as evidence, participants, and decisions evolve.

Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.