Skip to content
Designed Operational engine

Pulse

Workspace measurement — keeps documentation accurate (drift) and produces reports (campaigns).

Replaces manual metrics and stale docs

In the constellation

Pulse highlighted in the live map — hover or tap any node to explore.

01Why

Why it exists

The numbers in your documentation start drifting the moment anything changes — a count, a version, a metric that was right last month and quietly went wrong. Nobody notices until it bites, because re-checking every claim by hand never scales.

And the recurring questions about your own system — what’s current, what’s gone stale, what’s out of compliance — get answered by whoever happens to go digging, inconsistently and never on a schedule.

02What it is

What Pulse is

Pulse is a workspace measurement engine. It keeps the numeric claims in your documentation synchronized with the real source of truth, and it runs declarative measurement campaigns — defined questions, run on a cadence — that produce reports about the state of your workspace.

It never rewrites blindly: like a good refactor tool, it shows you exactly what will change before anything is written, and routes every edit through a safe, reviewable plan.

03Capabilities

What it does

Drift detection & repair

Bind a number in your docs to its real source, and have stale claims caught and corrected the moment that source changes.

Declarative campaigns

Define what to measure — population, dimensions, method — once, and have it run on schedule and emit a report every time.

Scheduled reporting

Produce up-to-date snapshots of workspace state — freshness, counts, compliance — automatically rather than on a manual scramble.

Review-before-write safety

Every correction is previewed as a plan you approve; nothing is rewritten without you seeing the exact change first.

04The cleavage

The line between judgment and machinery

AccelMars draws one hard line through every product: what an AI decides, and what runs deterministically. Pulse sits on the boundary — and keeps the two honest.

AI judgment

Where AI helps

  • Suggesting which prose numbers to track and to what source
  • Fuzzy classification of harder-to-measure claims
Deterministic

What stays deterministic

  • Scanning state and detecting drift
  • Running campaigns and writing reports
  • Applying every edit through a reviewable plan

Measurement and writeback are mechanical and auditable; AI only assists with the fuzzy edges, like proposing what a number should be bound to. The corrections themselves are precise.

05Connections

How it connects

Designed

Designed; not yet built — the architecture and two-mode model are specified, ahead of a first implementation.