IT Knowledge
Infrastructure
Transformation
SR. TECHNICAL WRITER / CONTENT STRATEGIST · ALLANA JACKSON · DeptZero
RETRIEVAL
DOCUMENTED
RECOVERED
ZERO CRED. INCIDENTS
FRAGMENTED KNOWLEDGE ACROSS A HYPER-GROWTH ENGINEERING ORG
DoorDash IT ran across six engineering teams — Incident Response, Security Engineering, Platform Engineering, SRE/DevOps, Engineering Enablement, and IT Operations — but its knowledge infrastructure had never scaled to match. Documentation had grown without governance, tooling, or ownership, leaving the organization operationally exposed in ways the engineering quality didn't reflect.
Documentation was spread across 25+ repositories — Confluence, Notion, Google Drive, and GitHub — with no unified taxonomy, no ownership model, and no freshness indicators. Runbooks were so scattered that on-call engineers reported a 12-minute average retrieval time during live incidents, four times the 3-minute platform target. Engineering hours bled into documentation search, tribal knowledge replaced documented process, and security gaps traced directly to missing or untested authentication guides.
What was needed was a structured audit and rebuild — one that would produce a deployable content architecture, a governance model that could survive team turnover, and a measurable reduction in operational friction.
A THREE-LAYER KNOWLEDGE INFRASTRUCTURE WITH EMBEDDED GOVERNANCE
The system was built in three layers: a consolidated content architecture that eliminated tool sprawl, a governance framework that assigned ownership and enforced review cadence, and a discovery layer that surfaced the right content at incident speed.
The architecture consolidated 25+ repositories from six teams into a single Confluence-anchored SSOT, governed by a taxonomy that assigned every content type a standardized template, a named owner, and a review trigger. Security-critical documentation — auth integration guides, credential rotation procedures — was rebuilt from scratch and validated end-to-end in a sandbox environment before publication.
AUDIT → ARCHITECTURE → BUILD → VALIDATE → GOVERN
The engagement ran across six phases: Discovery, Architecture, Content Build, Search Specification, QA, and Governance Handoff.
- Phase 1 — Knowledge Audit Conducted structured interviews with 6 engineering teams. Mapped 25+ repositories spanning Confluence, Notion, GitHub, and Google Drive. Scored each against a four-factor rubric: accuracy, ownership, freshness, and findability. Identified P0 gaps in incident response and security documentation.
- Phase 2 — Information Architecture Design Built a unified content taxonomy with standardized naming conventions, folder structure, and content type classifications. Resolved the naming fragmentation that had produced six or more variant titles for identical procedures across teams. Delivered an IA specification document as the governing reference.
- Phase 3 — Content Rebuild & Migration Rewrote P0 runbooks to a standardized template with consistent inputs, decision trees, escalation paths, and rollback steps. Sandbox-validated all Okta and SAML integration guides against a live test environment. Authored the Engineering Golden Path onboarding sequence to replace three competing "getting started" documents.
- Phase 4 — Search Parameter Specification & Validation Authored the engineering brief for five ML-informed semantic search parameters, defining each parameter's values, weighting rationale, and expected retrieval behavior. Partnered with a platform engineer who implemented the parameters against Confluence's ML ranking layer. Validated retrieval accuracy against a test query library drawn from real incident reports, iterating until the three-minute retrieval target was met.
- Phase 5 — QA & Accuracy Validation Ran a structured QA pass across all content in the new system. Tested runbooks in staging against live incident simulations. Validated credential documentation against security team sign-off. Achieved a 98% QA pass rate; zero credential incidents post-migration.
- Phase 6 — Governance Framework & Handoff Delivered an ownership assignment matrix covering 100% of content in the new system. Designed a 90-day review cadence with automated Confluence reminders. Authored a governance playbook covering the deprecation protocol, freshness scoring rubric, and escalation path for ownership gaps. Conducted a live walkthrough with Engineering Enablement as the internal steward.
MEASURABLE OPERATIONAL IMPACT ACROSS ENGINEERING AND SECURITY¹
The engagement delivered a governance infrastructure that outlasted the project itself. The ownership model ensured no document in the new system could go unmaintained. The 90-day review cadence was adopted as a standing operational practice by Engineering Enablement. The Golden Path onboarding sequence became the canonical reference for new engineering hires, replacing the three competing guides that had previously sent new engineers straight to IT support in their first week.
KNOWLEDGE INFRASTRUCTURE BLUEPRINT
Navigate the three panels — Discovery & Audit, Infrastructure Design, and Knowledge Governance & Impact — to review the full scope of the engagement: audit methodology, repository inventory, pain point matrix, IA taxonomy, semantic search parameter specification, modular runbook templates, content ownership model, and measured outcomes.