Case Study: Oculus
Enterprise eText Publishing & Digital Product Configuration at Scale
Background
In higher education digital publishing, an eText is far more than a digitized book. Modern digital courseware titles are interactive software products embedded with rich pedagogical tools, synchronized audiobooks, strict digital rights management (DRM) thresholds, accessibility adaptations, and deep integrations into higher-ed learning ecosystems like MyLab, Mastering, and Revel.
Historically at Pearson, configuring and deploying these digital titles was severely fragmented. Production and editorial teams had to manage eText assets across multiple incompatible authoring pipelines, while critical product settings (such as printing limits, offline study tools, text-to-speech audio streams, and packaging configurations) were handled through disparate backend scripts, manual database updates, and email chains. Setting up a new title or releasing a configuration update took weeks of engineering ticketing, introducing frequent configuration drift and heightening risk during critical back-to-school enrollment windows.
The Oculus initiative was launched to solve this operational bottleneck. As Product Manager, I led the roadmap, requirements architecture, and cross-functional delivery to transform Oculus into Pearson’s central enterprise management console for eTexts. We consolidated multi-source ingestion, declarative feature configuration, automated QA tracking, and multi-channel delivery into a single, high-performance platform.
Project Overview
Oculus was engineered to bridge the gap between upstream content creation and downstream student-facing delivery. The platform operates as the central control plane where editorial managers, production specialists, and quality assurance engineers configure, validate, and publish digital learning products. Key workstreams included:
- Unified Product Ledger: A centralized management interface featuring dual Table and Grid views, instant faceted search, and role-based permissions across publishing lifecycle states (Draft, In-Review, QA Approved, Live).
- Declarative Configuration Engine: Centralized controls for digital rights, interactive study tools (notebooks, flashcards, shared annotations), synchronized audio narration, and platform feature flags without code deployments.
- Multi-Channel Packaging Bridge: Automated bundling pipelines that package eText content either as standalone consumer digital books or as embedded interactive modules inside larger courseware suites.
- Automated QA & Jira Synchronization: Programmatic integration with Jira to log all configuration mutations, generate an immutable audit trail, and gate release deployments on automated health validations.
- Scalable Cloud Microservices: High-performance architecture built to withstand high concurrency and rapid query throughput during peak semester launch cycles.
Challenges Faced
1. Multi-Tenant Packaging & Distribution Complexity:
A single educational title often needed to be distributed in multiple formats: as an unbundled, standalone eText accessible via web and mobile apps, or as an embedded module integrated deeply into platform gradebooks and LMS containers. Harmonizing these divergent packaging requirements without creating duplicate product copies or divergent metadata schemas required an abstract, multi-tenant product definition model.
2. Granular Rights & Accessibility Governance:
Different academic disciplines, institutional contracts, and copyright agreements dictated distinct operational permissions. Some titles required strict page-printing quotas and DRM restrictions; others required open offline study downloads and synchronized text-to-speech audio engines for accessibility compliance. Managing these granular, title-level rules without complex manual database modifications was critical.
3. Operational Blind Spots & Configuration Drift:
Prior to Oculus, configuration modifications were made via ad-hoc scripts or undocumented database writes. When a title experienced a display issue or audio desynchronization in live classrooms, tracing which setting was modified and by whom was nearly impossible. We needed strict system-level traceability and automated audit logging.
4. High Concurrency During Peak Registration Surges:
Higher education publishing experiences massive traffic spikes during fall and spring back-to-school windows. The Oculus backend needed to maintain sub-second search and configuration speeds while simultaneously synchronizing large product manifests across distributed cloud environments.
Solutions and Execution
1. Centralized Product Ledger & Dual-View Experience:
To accommodate different operational workflows, we designed a responsive React.js and Redux interface offering both Table and Grid views. Production coordinators managing hundreds of titles could filter by project, discipline, lifecycle status, or publication date, while editorial specialists could inspect rich visual card previews of individual titles. Granular role-based access control (RBAC) via enterprise Single Sign-On (SSO) ensured that sensitive licensing and publishing actions were restricted to authorized personnel.
2. Declarative Product Configuration Matrix:
Rather than relying on engineering support to hot-patch settings, we created a declarative configuration engine that abstracted complex backend flags into an intuitive administrative dashboard:
| Product Configuration Dimension | Legacy Manual Process | Oculus Centralized Management | Operational & Pedagogical Impact |
|---|---|---|---|
| Interactive Study Tools | Hardcoded into delivery templates; required code rebuilds to modify features. | Declarative toggles for notebooks, flashcard sets, and collaborative instructor highlights. | Instructors tailor study feature sets to course needs; zero engineering handoffs required. |
| Audiobook & Text-to-Speech | Handled via disconnected media servers and manually verified audio pointers. | Integrated media configuration linking synchronized audio files to text slates with playback speed controls. | Guaranteed WCAG accessibility compliance across 100% of enabled catalog titles. |
| Print Quotas & DRM Rights | Configured through manual license server scripts and static database tables. | Granular rights sliders setting allowable print percentages, offline download limits, and session caps. | Strictly protected publisher copyright while fulfilling institutional student accommodations. |
| Courseware Bundling | Bespoke packaging routines for standalone vs. LMS-integrated titles. | Reusable product manifests establishing one-click publishing to standalone eText or platform bundles. | Eliminated duplicate data entry and prevented configuration mismatches across channels. |
3. Automated QA & Jira Workflow Synchronization:
Quality assurance in educational courseware is paramount. To eliminate manual verification delays, we engineered a bidirectional integration between Oculus and Jira. Every time a product configuration is modified or submitted for staging, Oculus automatically generates an auditable ticket, logs the payload delta, and triggers automated pre-release health checks.
Titles cannot advance to the Live state without programmatic verification and verified QA signoff recorded in the ticket history. This established an immutable, enterprise-grade audit trail that eliminated post-deployment finger-pointing and dramatically reduced production defects.
4. Modern Cloud Architecture & Telemetry:
To ensure stability, scalability, and long-term maintainability, we architected Oculus on a modern enterprise cloud stack:
- Backend Microservices: Developed with Java and Spring Boot for robust transaction management, data integrity, and seamless integration with legacy enterprise backends.
- Data Layer: MongoDB document store for dynamic, flexible product schema definitions that readily accommodate diverse title configurations without rigid table migrations.
- Containerized Infrastructure: Docker containers orchestrated on AWS Fargate, providing elastic auto-scaling that scales compute capacity up during seasonal enrollment surges and scales down to optimize costs.
- Performance Monitoring: Real-time APM telemetry via New Relic and AWS CloudWatch health-check endpoints, continuously monitoring API transaction response times and database connection pools.
Outcomes and Business Impact
Oculus revolutionized digital product operations across Pearson's higher education division, delivering quantifiable improvements in velocity, accuracy, and platform reliability:
- 60% Reduction in Configuration Turnaround: Replaced multi-week manual ticketing and engineering handoffs with a self-service console where production teams can configure and publish titles in hours.
- 100% Audit Compliance & Traceability: Integrated Jira logging created a comprehensive, tamper-proof record of every configuration edit, permission change, and publishing event across the catalog.
- High-Concurrency Scalability: Maintained sub-2.5 second API latency and 99% uptime throughout high-volume back-to-school surges, supporting thousands of concurrent instructor and student sessions.
- Eliminated Multi-Channel Configuration Drift: Standardized product manifest definitions ensured that standalone eTexts and bundled LMS courseware instances maintained identical content fidelity and rights governance.
- Broad Accessibility Adoption: Centralized audiobook and text-to-speech configuration enabled rapid rollout of accessibility compliance features across hundreds of previously unserviced titles.
Strategic Product Trade-offs
Declarative Standardization vs. Bespoke Feature Forks
During platform rollout, individual academic discipline teams frequently requested one-off custom features, such as bespoke highlighting colors, non-standard study tools, or specialized navigation layouts for specific flagship textbooks. Editorial managers argued that creating separate code forks would meet immediate author deadlines.
From a product standpoint, we resisted the temptation to create bespoke forks. Allowing one-off exceptions would have splintered the codebase, created unmaintainable regression testing cycles, and crippled downstream automated compilation.
Instead, we established a strict architectural boundary: every requested capability was evaluated and abstracted into a declarative configuration schema. If a feature had pedagogical merit, it was built as a reusable, toggleable parameter available to the entire catalog. While this required disciplined requirement analysis up front, it protected platform scalability and satisfied over 95% of customized discipline needs without incurring technical debt.
Lessons Learned
1. Configuration-as-Data Beats Bespoke Code:
Hardcoding product features into delivery scripts creates an operational bottleneck that slows down the entire business. Modeling product flags, study tools, and DRM limits as declarative data empowered non-technical business teams to move independently while keeping the engineering core clean.
2. Operational Empathy Drives Adoption:
Publishing production teams handle enormous title volumes under tight semester deadlines. Designing dual Table and Grid workflows with quick filters and bulk actions demonstrated an understanding of their daily friction points, transforming skeptical legacy users into platform champions.
3. Auditability Must Be Built-In, Not Bolted On:
Integrating Jira directly into the publishing pipeline from day one eliminated ambiguity. When every setting mutation is tied to an auditable event, QA teams spend their energy validating content rather than investigating who changed a setting.
4. Plan for Seasonal Concurrency Early:
Higher education traffic is intensely cyclical. Decoupling background compilation from real-time administrative querying and utilizing containerized AWS Fargate infrastructure ensured that seasonal registration rushes caused zero degradation in production performance.
Conclusion
The Oculus case study highlights the impact of bringing modern product management principles to enterprise internal tools. By replacing fragmented scripts, spreadsheet tracking, and manual handoffs with a unified, declarative configuration platform, the project drastically accelerated digital product velocity, eliminated configuration errors, and protected live classroom delivery.
Leading this initiative reinforced the value of treating internal operations platforms with the same rigor as customer-facing applications, aligning architecture with long-term business strategy to unlock operational agility at enterprise scale.