The platform
One explainable answer for every release.
Reloqui reconciles the signals scattered across your SCM, trackers, CI/CD, registries, deploy tooling and Terraform into trusted release truth — what shipped, where, and how confident we are, with the evidence to prove it.
How Reloqui works
From raw delivery signals to a publish decision you can defend — in four stages.
Ingest evidence
Twelve connectors stream typed events — merged PRs, issues, builds, artifacts, deployments, Terraform runs, incidents — via webhooks and polling.
Reconcile
An 11-stage, deterministic-first engine resolves the release window and attributes every signal to the right service and version.
Classify & score
Each item is classified (included, heuristic, excluded, uncertain, orphaned) and the release earns a confidence score with a full breakdown.
Decide & publish
Policy gates and N-of-M approvals decide the outcome; publishing freezes an immutable, auditable snapshot.
Built on six pillars
Reconciliation engine
Deterministic-first, heuristic-second attribution with clock-skew tolerance, backport detection and force-push handling — the moat that turns scattered signals into truth.
Evidence model
Every claim is a typed, stored event. Raw payloads are archived; normalized events drive the answer. Nothing is taken on faith.
Service catalog & graph
Model products, services, components and resources with a typed dependency graph, lineage and blast-radius analysis.
Governance
Confidence policies, N-of-M approvals, freeze windows and automatic suspension when late evidence collapses trust.
Audience notes
Executive, engineering, operations, product and compliance narratives — generated from the same evidence base, every claim traceable.
360° views & insights
Release, Service and Product 360, live dashboards, drift detection and full-text search across your delivery estate.
Confidence you can defend
Every release carries a 0–1 confidence score and a high / medium / low tier, built from weighted evidence components — tag anchors, linked PRs, artifacts, deployments, green builds — minus deductions for orphans, force-pushes and rollbacks. You see the exact breakdown, not a black box.
- ✓ Coverage-gap handling — never penalized for evidence you can't produce
- ✓ Exceptions surfaced, not hidden — orphans, uncertain items, force-pushes
- ✓ Explainable reason on every classified item
Replace the release meeting with a link.
Connect a repository — or generate a synthetic evidence trail — and publish your first evidence-backed release in under ten minutes.