Project Name
Fragmented Government Docs Unified on Backstage, Onboarding From 6 Weeks to 3 Days
A large North American government agency managing OCI-hosted services across multiple departments had institutional knowledge scattered across Confluence, SharePoint, wikis, email threads, and individual laptops. New engineers took 4-6 weeks to reach productivity. SREs lost 15-30 minutes per incident hunting runbooks. Documentation lived outside source control, owned by no one, drifting away from reality as services evolved – with no defensible audit trail for regulatory compliance. Applying its AI-First approach, Ksolves deployed Backstage TechDocs with a Dagger-automated docs-as-code pipeline – collapsing five knowledge silos into one searchable portal, owned by teams, versioned in Git, always in sync with live OCI services.
- Institutional Knowledge Fragmented Across Five Systems: Runbooks, architecture decisions, and operational procedures split across Confluence, SharePoint, wikis, email chains, and individual laptops - no single authoritative source, frequent version conflicts.
- New Engineer Onboarding Taking 4-6 Weeks: Without a centralised documentation portal, new engineers spent weeks in knowledge-transfer meetings and hunting runbooks - a productivity lag that multiplied with every hire.
- Incident Runbook Discovery Extending MTTR: SREs lost 15-30 minutes per incident locating the correct runbook across disconnected systems - directly violating SLA commitments and extending mean time to resolution on every OCI service incident.
- Documentation Perpetually Out of Date: Docs lived outside source control and were owned by individuals. As services evolved, documentation drifted - engineers stopped consulting docs because they had learned they were unreliable.
- No Service Catalog or Ownership Registry: No registry mapped OCI services to owners, dependencies, or runbooks. Determining who to contact for a specific OKE workload during an incident required tribal knowledge or Slack threads.
- Regulatory Auditability Gap: Government compliance requirements mandate documented operational procedures for all production services. With documentation scattered and unversioned, the agency had no defensible audit trail proving runbooks were current, reviewed, and owned.
Ksolves deployed Backstage TechDocs as the unified documentation portal and Dagger (Python SDK) as the automated publish pipeline. One principle: every document lives in Markdown alongside service code in GitLab, built on every commit, published to OCI Object Storage, and surfaced in Backstage linked to its service catalog entry. No doc outside version control. No runbook owned by a person.
- Backstage TechDocs Unified Portal: All runbooks, API docs, and architecture decisions indexed and searchable in Backstage TechDocs, linked to their service catalog entry. Any runbook discoverable in under 30 seconds through a single search interface regardless of which department owns the service.
- Dagger Python SDK Docs Pipeline: Complete docs pipeline - MkDocs build, link validation, HTML generation, and publish to OCI Object Storage - triggered on every Git commit. Backstage always serves documentation at source-code parity, drift eliminated automatically.
- Backstage Service Catalog With Full OCI Coverage: Every OCI service - OKE workloads, OCI Functions, Oracle DB instances - registered with owner, dependencies, environment status, and runbook link. Tribal knowledge replaced with a single queryable registry.
- Docs-as-Code Migration From Legacy Systems: Confluence pages, SharePoint documents, and wiki entries migrated to Markdown in GitLab repos alongside service codebases - team-level ownership, Git versioning, and peer review through the same merge-request workflow used for code.
- OCI DevOps Governed Pipeline Runs: All Dagger documentation pipelines orchestrated through OCI DevOps - every documentation state traceable to a specific Git commit, merge request, and pipeline run for regulatory compliance.
Technology Stack
| Category | Technology |
|---|---|
| Platform | Backstage + TechDocs |
| CI/CD | Dagger (Python SDK) |
| Storage | OCI Object Storage |
| Source Control | Git / GitLab |
| Compute | OCI Kubernetes Engine (OKE) |
| Infrastructure | OCI DevOps |
- Onboarding Cut From 6 Weeks to Under 3 Days: All runbooks and architecture decisions discoverable in Backstage in under 30 seconds - no knowledge-transfer meetings required, productive onboarding window collapsed from weeks to days.
- Runbook Discovery Cut to Under 60 Seconds: Backstage global search surfaces any runbook by service name, symptom, or keyword in under 60 seconds - MTTR directly reduced and per-incident documentation hunts eliminated.
- Documentation Drift Eliminated: Dagger pipelines rebuild and republish on every Git commit - Backstage always serves documentation at source-code parity with no manual maintenance required.
- First Complete OCI Service Ownership Registry: Backstage service catalog provides real-time ownership, dependency maps, and runbook links for every OKE workload, OCI Function, and Oracle DB - incident response starts with a catalog lookup, not a Slack thread.
- Regulatory Audit Trail Established: Every documentation state traceable to a Git commit, merge request, and OCI DevOps pipeline run - the agency can now prove all runbooks are current, owned, peer-reviewed, and version-controlled for compliance auditors.
A government agency losing institutional knowledge across five disconnected systems, with 6-week engineer onboarding and 30-minute per-incident runbook hunts, was unified on a single searchable portal through Ksolves DevOps consulting services. Backstage TechDocs and Dagger docs-as-code pipelines collapsed five silos into one – onboarding dropped to under 3 days, runbook discovery cut to under 60 seconds, documentation drift eliminated, and a regulatory-grade audit trail established for every production service. Every runbook is now versioned, team-owned, peer-reviewed, and traceable to a Git commit.
Losing Institutional Knowledge Every Time an Engineer Leaves?
