Token, Optimisation, Policy, Auditability and Security.
TOPAS is the token-control and inference-efficiency engine within HEX 165. It governs how AI workloads consume context, spend budget, select models and create value.
TOPAS applies policy, sovereignty, accuracy and cost controls before the model call, then records every material decision to the immutable audit trail.
Large-context AI creates capability, but every additional token can increase cost, latency and governance exposure.
TOPAS measures cost per completed task, not just cost per token. It controls model routing, context size, cache efficiency, repeated payloads, retries and human review, helping organisations prevent AI cost overspend before it happens.
Select the lowest-cost compliant model path after policy, residency, SLA and accuracy checks.
Optimise repeated and structured context through cache alignment and Group of Tokens (GOT).
Validate compressed outputs against full-context reference results, with original source material retained in the Restore Vault.
Compare AI execution cost with equivalent human effort, using approved role, skill and rate benchmarks.
Record every material route, cost, compression, restore, approval and value decision to the immutable audit trail.
A governed AI add-in for Microsoft Word. It puts policy, model routing, token control, cost visibility and audit evidence inside document review and drafting workflows.
A developer-facing sidecar for applications, agents and delivery pipelines, bringing TOPAS policy, routing, budgets, GOT optimisation and audit telemetry into AI-enabled software.
TOPAS works alongside the Governance, Compliance and STPM Engines as part of one control plane for regulated agentic AI.