Pete Ratkevich Product Manager

Case Study: Unified Dashboard

Consolidating a Fragmented Content Supply Chain into a Single Enterprise Platform

Unified Dashboard

Background

Pearson's digital content production teams were operating across six disconnected tools (a standalone narrative authoring canvas, three separate assessment engines, a siloed media asset library, and a configuration tool) with no shared state, no unified versioning, and no central visibility into where any given title stood in its lifecycle. The result was a content supply chain held together by manual handoffs, tribal knowledge, and spreadsheets.

Producing a new educational title from development to deployment took 6–12 months. In-flight corrections, even simple ones, required engineering escalations that could take weeks. The platform fragmentation wasn't just a workflow inconvenience; it was a strategic ceiling on how fast the business could move.

The Unified Dashboard was the initiative to fix that. As Product Manager, I led the roadmap, requirements definition, and cross-functional execution to consolidate those six tools into a single enterprise publishing platform, cutting title velocity from months to weeks and enabling content updates to go live within 24 hours.

6–12 w
Product Velocity
Dev-to-deployment compressed from 6–12 months down to weeks.
< 24h
In-Flight Updates
Enabled same-day publishing updates for live educational titles.
6 → 1
Tool Consolidation
Unified authoring, 3 assessment tools, media DAM, & config into 1 console.
< 3.0s
API Concurrency
Peak search and transaction latency stabilized across microservices.

Project Overview

The project scope spanned the full content production lifecycle: from initial title setup and structured authoring through media integration, assessment validation, and multi-channel distribution. Key workstreams included:


Challenges Faced

1. Complex Legacy Tool Integrations:

The project required integrating several independent enterprise systems, each built with its own proprietary data schema, validation states, and unique identifiers. For example, digital asset repositories required strict image optimization protocols, while assessment engines operated under distinct unique identifiers (URNs) aligned with separate gradebook configurations. Harmonizing these systems into an intuitive authoring picker without incurring massive technical debt demanded careful middleware sequencing and robust API contracts.

2. Scalability and High Concurrency:

Given the platform's central enterprise role, the Unified Dashboard needed to handle massive datasets, complex relational content trees, and high user concurrency during back-to-school release peaks. Initial load testing revealed latency spikes during heavy manifest generation and multi-cluster distribution routines that required architectural remediation.

3. User Experience (UX) Across Diverse Personas:

The platform served a wide spectrum of users, ranging from non-technical editorial authors and pedagogical designers to specialized production engineers and external phase vendors. Balancing intuitive ease-of-use with advanced structural controls was essential: non-technical users needed simplicity, while power users required deep access to indexing, MathML notation, and metadata configurations.

4. Abstracting Content Structure from Presentation:

A fundamental product challenge was shifting author mindsets away from traditional WYSIWYG ("What You See Is What You Get") styling toward a strict semantic structure model. Authors were accustomed to hardcoding fonts and layouts. The platform needed to decouple content meaning from visual display so that a single raw content block could dynamically reformat across desktop web browsers, mobile apps, ePub readers, and print-on-demand masters.

5. In-Flight Courseware Protection:

Updating titles while classes were actively in session posed severe operational risks. Pushing unvetted structural edits to live courseware could desynchronize active student gradebooks, break submission trees, or drop active course containers. We needed clear system guardrails to govern in-flight updates safely.


Solutions and Execution

1. Combining 6 Disconnected Tools into a Unified Dashboard:

Rather than requiring authors and production staff to constantly switch between standalone tools, we orchestrated the entire digital content supply chain into a unified, single pane of glass:

2. Modern Technology Stack:

We selected a modern, scalable tech stack to support the unified ecosystem and ensure rapid deployment:

3. Cross-Functional Discovery & Iterative Delivery:

Integrating six legacy systems into a single platform meant that requirements were evolving as we discovered constraints. We structured delivery in bi-weekly sprints with explicit integration milestones, sequencing the most technically risky work first (the federated API contracts and assessment bridge) to surface blockers early in the roadmap before they became blockers at scale.

Adoption risk was managed through continuous co-design: editorial teams, production engineers, academic authors, and external vendors all participated in prototype reviews and usability sessions before development was finalized. This compressed rework cycles and created internal advocates who helped drive adoption on launch.

4. Operational Governance & In-Flight Update Framework:

To safeguard active classrooms and live student gradebooks during subsequent title revisions, we established a clear Major vs. Minor Update Governance Framework. The system programmatically calculates the blast radius of any modification and enforces automated safety controls:

Update Type Content Modification Scope System Enforcement & Action Risk Profile & Impact
Minor Update Inline narrative typos, clarifying question prompts, or media asset pointer replacements. Hot-patches directly within the active container block; updates asset pointers via the DAM picker. Low Risk: Safely propagates into live courses without altering student submission records or database keys.
Major Update Replacing assessment containers, modifying scoring logic, or restructuring the Table of Contents. Blocks direct live edits. Enforces a full project clone and automatically generates a new distinct Product ID. Critical Risk Mitigated: Absolute branch isolation guarantees active student gradebooks and course histories are never corrupted.

5. Shifting Quality Left: The Automated Preflight Engine:

Instead of treating quality assurance as an end-of-pipeline manual gate, we engineered an automated Preflight Quality Engine built directly into the authoring canvas.

Before a title can be staged or deployed, the system programmatically parses the project structure, verifying that media containers have valid assets with alt-text, questions map to active Learning Objectives, and hyperlinks are intact. If discrepancies are found, the pipeline halts and generates an actionable remediation sheet with deep links to the exact canvas location. This proactive validation cleared 75% of structural defects before staging, eliminating late-stage engineering hotfixes.

Distribution Loop Endpoint Peak Latency Optimized Latency Architectural Root Cause & Strategic PM Mitigation
Publishing Manifest Generation 22.43s 3.80s Mitigation: Re-architected data fetching into parallel batch requests segmented by date windows, relieving database connection pools.
Multi-Cluster Distribution Loop 22.10s 3.20s Mitigation: Converted synchronous cluster validation loops into an asynchronous polling worker model, decoupling user sessions.
Project Status Synchronization 14.76s 2.10s Mitigation: Introduced an in-memory Redis caching layer for project manifests, bypassing repetitive full-table database scans.

Outcomes and Business Impact

The Unified Dashboard transformed Pearson's digital content publishing from a fragmented, high-friction operation into a modern, scalable ecosystem:


Strategic Product Trade-offs

Resisting the WYSIWYG Trap for Long-Term Scalability

During discovery, editorial users requested an unrestricted WYSIWYG editor: full control over fonts, sizing, and layout. It was a reasonable ask from people accustomed to Word and InDesign.

We pushed back. Permitting inline visual overrides would have injected unstructured formatting markup directly into database rows, destroying the multi-channel portability that was central to the platform's value proposition and breaking responsive rendering on mobile, ePub, and print targets.

Instead, we enforced a clear system boundary: authors define pure semantic structure (headers, narrative blocks, callouts, formulas), and visual styling is injected dynamically during downstream compilation per target format. The tradeoff required additional onboarding investment, but it future-proofed Pearson's content library and made single-source multi-channel publishing viable at scale.


Lessons Learned

1. Early users become your best advocates:

Migration from familiar legacy tools carries real adoption risk. Involving editorial and production teams in design workshops from the start, before the build, turned potential resisters into advocates who championed the platform internally on launch.

2. Sequence risk, don't just manage it:

On a multi-system integration, the most technically uncertain work should come first. By front-loading the federated API contracts and assessment bridge in the roadmap, we exposed cross-system compatibility issues early, when they were still cheap to fix.

3. System guardrails beat policy:

Telling teams "be careful updating live courses" is not a governance strategy. Encoding blast-radius detection into the platform itself, blocking certain actions and enforcing branch isolation automatically, removed the risk from human hands entirely.

4. Platform thinking pays compounding dividends:

Every decision made to decouple content structure from presentation, to build shared API contracts, and to centralize configuration management created headroom for future capabilities. The Unified Dashboard wasn't just a tooling improvement. It was a content platform that Pearson could build on.


Conclusion

Six tools. One platform. Development cycles measured in weeks, not months. The Unified Dashboard is the kind of PM work I find most meaningful: a complex systems problem where product strategy, technical depth, and cross-functional leadership all have to fire together to get a real outcome.

This project required holding a clear product vision across a long, ambiguous integration, managing stakeholders ranging from non-technical editorial authors to senior engineering architects, while making principled tradeoffs that prioritized long-term platform health over short-term requests. The result was a publishing operation that moved faster, broke less, and opened up capabilities that simply weren't possible before.