Trust Engine Overview

The Trust Engine runs five evaluation pillars in parallel against every model output and returns a single, policy-weighted verdict.

Architecture

The Trust Engine is a stateless evaluation service. For each request it:

1

Tokenizes and preprocesses the prompt + response pair.

2

Dispatches all five pillars concurrently under a shared latency budget.

3

Aggregates pillar scores using your policy weights.

4

Resolves the verdict from your thresholds (plus any hard-block flags).

5

Returns the result and emits an audit record.

The pillars

Safety & Toxicity

Harmful, violent, harassing, self-harm content.

Hallucination

Fabricated facts and citations.

Bias & Fairness

Stereotyping and discriminatory framing.

Prompt Security

Injection and jailbreak detection.

Compliance & PII

Regulated data and personal information.

Scoring model

Each pillar emits a score in [0,1] and a confidence value. The overall trust score is the weighted aggregate of pillar scores. Because weighting is deterministic, the same input and policy version always produce the same verdict — essential for reproducible audits.

Latency tiers

You choose an SLA tier per request via the x-veldrix-sla-tier header (or your plan default). Per-pillar work is generous, but the total budget is the hard deadline:

Tierp95 budgetUse for
REALTIME≤ 200 msInteractive, user-blocking calls.
STANDARD≤ 500 msDefault for most traffic.
BACKGROUNDuncappedBatch / fire-and-forget logging.
Run evaluations in background mode and the tier never touches your user-facing latency at all.

Grounded evaluation

Pass source context (retrieved documents, system instructions) in the context field and pillars evaluate the response against that ground truth — sharply reducing hallucination false-positives versus context-free checking.

Was this page helpful?