Enterprise Deployment Orchestration Platform

A program framing delivery modernization as a shared platform capability for engineering teams.

CI/CDRelease OrchestrationPlatformDevOps

Organization: Qualys
Role: Technical Program Manager
Industry: Cybersecurity Enterprise software
Duration: 2 quarters

Objective

Enterprise Deployment Orchestration Platform is an enterprise deployment orchestration platform designed to replace fragmented pod-wise Jenkins and legacy handoff workflows. After a focused Q2 2026 delivery cycle, the platform achieved a Conditional Go for non-FedRAMP microservice production deployments, with VM-based workloads and FedRAMP scope deferred to subsequent phases.

Business Challenge

Prior to Enterprise Deployment Orchestration Platform, deployment workflows at Qualys were fragmented and operationally expensive:

No standardized deployment workflow — Each pod and environment used Jenkins jobs or manual runbooks with inconsistent steps.
Manual coordination — Configuration sync, dependency checks, and deployment handoffs required heavy SRE and EngOps coordination.
Weak version binding — Application and configuration versions were not strongly coupled, increasing drift and rollback risk.
No centralized dependency registry — Teams could not automatically validate whether upstream services were healthy before deploying.
Limited multi-pod support — Bulk or coordinated releases across multiple pods were difficult to orchestrate.
Poor deployment visibility — Leadership, SRE, and engineering lacked a single pane of glass for deployment status, history, and health.
Inadequate auditability — Deployment traceability and compliance reporting were hard to reconstruct.
These gaps increased lead times, error rates, and operational risk as the engineering portfolio scaled. A unified platform was needed to standardize deployments while preserving safety and auditability.

Context

Implement Continuous Delivery (CD) as the enterprise-standard deployment orchestration platform for standardized, scalable, observable, and dependency-aware deployments across engineering and production pods.

Responsibilities

Own end-to-end program governance for Enterprise Deployment Orchestration Platform, including charter, roadmap, milestones, and delivery timeline.
Coordinate cross-functional stakeholders across engineering, SRE, DSRE, security, and leadership, and drive issues to closure through the right owners.
Manage risk, track production blockers, and ensure all go/no-go criteria are met before launch.
Facilitate security and compliance approvals (ARB, SecOps, QUASAR, ProdSec) while preventing FedRAMP scope from blocking the non-FedRAMP rollout.
Oversee operational readiness, runbooks, observability, rollback planning, and SRE sign-off.
Produce decision artifacts such as the go/no-go checklist, leadership communications, and case study, and present clear recommendations with documented sign-off.

Execution

The program was delivered in three phases.

Phase 1 — Initiation & Development (Feb–Apr 2026)

Program kickoff and stakeholder alignment
MVP backlog freeze and Jira epic/story creation
Pilot application selection
Core feature development: UI/backend, Config Watcher, Quad Agent, dependency graph, RBAC, LDAP, scheduling, and analytics

Phase 2 — Engineering Rollout & Validation (May–June 2026)

Dry-run environment setup on engineering pods
Pilot app onboarding and deployment on selected pods
Broader engineering pod rollout for microservices and VM-based services
SRE and EngOps validation of upgrade scenarios, dependency management, rollback, and parallel deployment
ARB, SecOps, QUASAR, GRC, SIEM, and Platform Security clearances obtained

Phase 3 — Production Readiness & Go/No-Go (July 2026)

Platform deployed on Ops-Staging
Microservice deployments validated on Ops-Staging
Production infrastructure readiness planning initiated by DSRE
Capacity planning reviewed: 80 concurrent deployments tested on a 16-CPU / 32-GB control node
Final leadership Go/No-Go review held

Outcomes

Pilot feedback positive
Core platform operationally ready
Microservice deployments validated
Observability and resilience in place
Operational runbooks available

Lessons Learned

Separate compliance tracks early. FedRAMP/SCR approvals move on a different cadence than engineering delivery. Ring-fencing scope allowed non-FedRAMP value to ship without being blocked.
VM and microservice deployments need distinct validation tracks. VM onboarding revealed unique blockers that would have delayed the entire program if bundled with microservices.
Human review of AI-generated configs is non-negotiable. Leadership explicitly required the two-person SRE review process to be documented, signed, and audited.
Capacity and DBRE readiness are first-class deliverables. Treating infrastructure readiness as a side activity creates last-minute risk; dedicated owners and tickets are essential.
Pilots surface integration issues better than synthetic tests. Real application pilots uncovered practical concerns that lab tests did not.
Operational runbooks and RACI are part of the product. A technically ready platform is not production-ready without documented support, escalation, and rollback procedures.
Keep leadership decision records explicit. The decision matrix and sign-off table provided clarity and defensibility for the Conditional Go call.

Objective

Business Challenge

Prior to Enterprise Deployment Orchestration Platform, deployment workflows at Qualys were fragmented and operationally expensive:

No standardized deployment workflow — Each pod and environment used Jenkins jobs or manual runbooks with inconsistent steps.

Manual coordination — Configuration sync, dependency checks, and deployment handoffs required heavy SRE and EngOps coordination.

Weak version binding — Application and configuration versions were not strongly coupled, increasing drift and rollback risk.

No centralized dependency registry — Teams could not automatically validate whether upstream services were healthy before deploying.

Limited multi-pod support — Bulk or coordinated releases across multiple pods were difficult to orchestrate.

Poor deployment visibility — Leadership, SRE, and engineering lacked a single pane of glass for deployment status, history, and health.

Inadequate auditability — Deployment traceability and compliance reporting were hard to reconstruct.

These gaps increased lead times, error rates, and operational risk as the engineering portfolio scaled. A unified platform was needed to standardize deployments while preserving safety and auditability.

Responsibilities

Own end-to-end program governance for Enterprise Deployment Orchestration Platform, including charter, roadmap, milestones, and delivery timeline.

Coordinate cross-functional stakeholders across engineering, SRE, DSRE, security, and leadership, and drive issues to closure through the right owners.

Manage risk, track production blockers, and ensure all go/no-go criteria are met before launch.

Facilitate security and compliance approvals (ARB, SecOps, QUASAR, ProdSec) while preventing FedRAMP scope from blocking the non-FedRAMP rollout.

Oversee operational readiness, runbooks, observability, rollback planning, and SRE sign-off.

Produce decision artifacts such as the go/no-go checklist, leadership communications, and case study, and present clear recommendations with documented sign-off.

Execution

The program was delivered in three phases.

Phase 1 — Initiation & Development (Feb–Apr 2026)

Program kickoff and stakeholder alignment
MVP backlog freeze and Jira epic/story creation
Pilot application selection
Core feature development: UI/backend, Config Watcher, Quad Agent, dependency graph, RBAC, LDAP, scheduling, and analytics

Phase 2 — Engineering Rollout & Validation (May–June 2026)

Dry-run environment setup on engineering pods
Pilot app onboarding and deployment on selected pods
Broader engineering pod rollout for microservices and VM-based services
SRE and EngOps validation of upgrade scenarios, dependency management, rollback, and parallel deployment
ARB, SecOps, QUASAR, GRC, SIEM, and Platform Security clearances obtained

Phase 3 — Production Readiness & Go/No-Go (July 2026)

Platform deployed on Ops-Staging
Microservice deployments validated on Ops-Staging
Production infrastructure readiness planning initiated by DSRE
Capacity planning reviewed: 80 concurrent deployments tested on a 16-CPU / 32-GB control node
Final leadership Go/No-Go review held

Lessons Learned

Separate compliance tracks early. FedRAMP/SCR approvals move on a different cadence than engineering delivery. Ring-fencing scope allowed non-FedRAMP value to ship without being blocked.

VM and microservice deployments need distinct validation tracks. VM onboarding revealed unique blockers that would have delayed the entire program if bundled with microservices.

Human review of AI-generated configs is non-negotiable. Leadership explicitly required the two-person SRE review process to be documented, signed, and audited.

Capacity and DBRE readiness are first-class deliverables. Treating infrastructure readiness as a side activity creates last-minute risk; dedicated owners and tickets are essential.

Pilots surface integration issues better than synthetic tests. Real application pilots uncovered practical concerns that lab tests did not.

Operational runbooks and RACI are part of the product. A technically ready platform is not production-ready without documented support, escalation, and rollback procedures.

Keep leadership decision records explicit. The decision matrix and sign-off table provided clarity and defensibility for the Conditional Go call.