sdlc.works · orchestration for the whole lifecycle

Your engineers ship code 3 days a week. You pay for five.

The other two go to environments that drifted, certs that expired, artifacts nobody owns, and audits that eat a quarter. sdlc.works runs that work as a platform obligation — provisioned, rotated, witnessed, and reconciled without a ticket.

Calculate what toil costs you See a real week

$ curl -fsSL sdlc.works/install | sh  ·  running in your cloud, not ours

platform activity — acme-corp / this morninghandled without a human
Three weeks that actually happen

The same incident, run twice.

Not hypotheticals — the recurring failure modes every engineering org pays for. Left column is the version your team lived last quarter. Right column is the platform's version.

The Saturday 2 a.m. certificate

incident · availability
How it went

The wildcard cert on the API gateway expired at 02:04. Checkout returned TLS errors for 4 hours. Three engineers paged; the one who knew the renewal runbook was on holiday. Postmortem action item: "add calendar reminder."

~40 eng-hours + 4h revenue impact + 1 ruined weekend
With sdlc.works

Certificate lifecycle is a declared integration on the branch. Renewal ran 30 days before expiry, the new cert was attested into the ledger, the rollout verified against the URL register. The on-call phone never rang.

0 eng-hours · one ledger entry · nobody paged

"Works on staging, dies in prod"

drift · lost sprint
How it went

A payments bug reproduced nowhere but production. Two engineers spent 2.5 days diffing environments by hand before finding it: staging had a feature flag and an env var that prod didn't. Neither change had an owner or a record.

~40 eng-hours + a slipped release + eroded trust in staging
With sdlc.works

Both environments are declared shapes of the same branch. The unrecorded env var showed up as drift within 41 seconds of appearing, with the diff, the source, and a one-click reconcile. The "bug hunt" is a log line.

~5 minutes · full diff history · staging you can trust again

The SOC 2 evidence quarter

audit · compliance
How it went

The auditor asked who deployed what, when, from which commit, with whose approval — for twelve months. A senior engineer spent 3 weeks screenshotting CI runs, exporting chat threads, and reconstructing releases from tribal memory.

~120 eng-hours of your most expensive people, every year
With sdlc.works

Every build, deployment, and provisioning event was attested at the moment it happened — commit, artifact hash, actor, target, timestamp — into an append-only ledger. Evidence for the full period is one export.

one afternoon · zero reconstruction · auditors love it
Who signs, who ships, who sleeps

Three seats at the table. One decision.

A platform purchase touches the person who approves it, the people who build on it, and the people who currently do this work by hand. Here's the deal each of them gets.

You're paying senior salaries for work a platform should own.

  • Recover capacity without hiring. Toil reclaimed across a 40-engineer org is roughly the output of 5–8 additional engineers — at a fraction of one loaded salary.
  • Make audits a line item, not a quarter. SOC 2 / ISO evidence exports from the attestation ledger replace weeks of manual reconstruction, every cycle.
  • Cut incident exposure you can point to. Expired credentials, drifted configs, and unowned cron jobs — the top recurring outage causes — become platform-enforced policy.
  • One system of record for "what is running where." Board question answered in one query, not a week of Slack archaeology.
~30%
of eng payroll typically spent on ops toil — adjustable in the calculator below
1 afternoon
to produce 12 months of deployment evidence for an auditor
0 headcount
added to run it — it runs in your cloud, declared once
The number for the budget meeting

What does toil cost you, specifically?

Drag the sliders to your org. The assumptions are printed, not hidden — take the math to finance.

assumptions: 46 working weeks/yr · sdlc.works reclaims a conservative 60% of declared-scope toil (env provisioning, drift, cert/secret lifecycle, artifact hygiene, audit evidence, scheduled-job upkeep) · numbers are illustrative — replace with your own time-tracking data for a board-grade figure
$1.30M
engineering payroll recovered / year
13,200
eng-hours / year back to shipping
7.2
equivalent full-time engineers, without hiring
Scope of ownership

The work it takes off your org, item by item.

No category labels, no adjectives. A concrete list of what moves from someone's Tuesday to the platform's job description.

Lifecycle work
How it's done today
With sdlc.works
Environmentsper-branch, per-PR
A ticket to the platform team; 1–3 days of waiting; a shared staging env everyone steps on.
Declared once per branch. Provisioned on PR open, reconciled on drift, destroyed on merge.
Certificates & secretsrotation, expiry, ownership
A spreadsheet of expiry dates, a calendar reminder, and an outage the year someone leaves.
Lifecycle policy as declaration. Rotated ahead of expiry, attested on rotation, alert only on failure.
Artifacts & registriesbuilds, images, retention
Registries filling with untagged images; nobody sure which build is actually in prod.
Every artifact witnessed — hash, source commit, deploy targets — with retention enforced, not hoped for.
Deploy & release recordswho, what, when, where
Reconstructed from CI logs and chat threads when something breaks or an auditor asks.
Append-only deployment ledger. One query answers "what is running where and since when."
Scheduled jobscrons, syncs, batch runs
Crontabs on boxes nobody logs into; overlapping runs discovered via duplicate invoices.
Schedules declared on the branch — overlap-guarded, monitored, with every run in the run ledger.
Third-party integrationsauth, webhooks, re-consent
Webhooks that silently stop firing; OAuth tokens that expire on a Friday.
Integration health as managed state. Re-auth surfaced before failure, not after the data gap.
Audit & compliance evidenceSOC 2, ISO 27001, customer DDQs
A senior engineer's quarter, annually, assembling screenshots.
Evidence is a byproduct. The attestation ledger exports the full trail in one afternoon.

Run the calculator. Then run the pilot.

Two weeks, one team, your cloud. If the ledger doesn't pay for itself, walk away with the data.

Start a two-week pilot Get the ROI worksheet