Case study — PulseCheck · pulsecheck.dataduke.net
The platform that connects the dots
PulseCheck is a real-time compliance and engineering-governance platform we built. It turns raw data from Jira, GitHub, GitLab, and CI/CD pipelines into scored compliance results, financial cost visibility, and AI-powered fix recommendations — no custom integrations required.
Eight tickets are marked Done in Jira. Three of their pull requests are still failing CI. Nobody has noticed, because the tools don't talk to each other.
PulseCheck detects the mismatch, opens a critical violation linking the exact tickets and pull requests, and the AI Coach drafts a rule so the check runs automatically every sprint from now on.
The data was always there. Now something connects the dots.
The problem
Everyone has the data. Nobody has the picture.
Engineering teams produce enormous amounts of process data. It sits siloed across Jira, GitHub, CI pipelines, and chat — and nothing connects it:
- Compliance drift happens silently between sprints — by the time it shows up, the cost is already baked in.
- Overdue tickets have a price tag that nobody calculates until the quarterly review.
- Working agreements are documented, then forgotten — sprint reviews still run on gut feel.
- Releases spanning several teams have no shared gate — one team's open critical bug becomes everyone's hotfix.
- Auditors ask for evidence that was never collected.
So we built the platform that connects the dots — and keeps score.
What we built
The guided tour — from raw data to a defensible score
PulseCheck syncs the tools you already use continuously, evaluates your own rules against live data, and turns the result into scores, costs, and alerts.
Rules your team can actually write
Compliance rules are written in JQL, YAML, or plain English — the AI turns a sentence into an executable rule. Template packs install in one click: DORA metrics, incident management, QA, GDPR, and the EU AI Act (Reg. 2024/1689), with violation messages that cite the relevant Article.
A compliance score you can defend
Every domain gets its own compliance percentage with a historical trend. Point-in-time snapshots stand as audit evidence, and every violation links straight to the exact ticket or pull request that caused it.
What a late ticket actually costs
Hourly rates are bound to team-member roles, so overrun is computed from real dates: (close date − deadline) × rate. Tickets are ranked by financial impact, and when cost drift breaches a threshold for consecutive sprints, an alert lands in Slack or Teams.
Engineering health — measured, not felt
DORA metrics are calculated automatically from CI/CD data: deployment frequency, lead time for changes, change failure rate, and time to recovery. No manual input, no spreadsheets.
Releases gated by facts
Multi-team releases are tracked on one board: each team must clear its compliance gate (say, ≥ 85%) before the release window opens. The approval workflow shows exactly what's blocking and who owns it.
An assistant that knows your numbers
The AI Coach knows the current score, violations, DORA metrics, and cost overruns — and through the MCP connector, Claude or Cursor can query your organization's live data too. Personal data is scrubbed before any AI call.
Trust & security
Built to pass its own audit
GDPR-compliant by design
EU data residency, personal data scrubbed before any AI call, no customer data used for model training, and full export on request.
Encrypted and isolated
Integration credentials and API keys are encrypted at rest (AES-256-GCM), with database-enforced row-level security on every table.
Audit-ready
Security audit trail, activity logging, and point-in-time compliance snapshots that stand as evidence when the auditor arrives.
No lock-in
Rules are portable JQL and YAML, and data export is available on every paid plan. If PulseCheck goes, your rules and history go with you.
PulseCheck is in production, with beta pilots running in the DACH region. We'd rather demo a live dashboard than a slide deck — ask us for a walkthrough.
Curious what your process would score?
See PulseCheck live, or talk to us about the compliance rules your team keeps in a wiki nobody reads. We'll show you what they look like as running checks.