From Nine Systems to One Search: Building an Integrated Product Knowledge Taxonomy for Pharma Manufacturing.

About the Client

Company AC is a global pharmaceutical manufacturer. Today, Company AC sells medicines in more than 130 countries. Within its manufacturing organization, the pharmaceutical technology group supports every established small-molecule product, with subject-matter expertise spanning API, drug product, manufacturing, equipment, analytical science, statistics, chemometrics, devices, packaging, and validation.

Industry Context

Large pharmaceutical manufacturers routinely accumulate product knowledge across a decade or more of separate systems, each built for a specific regulatory or functional purpose rather than for enterprise-wide findability. That fragmentation carries real weight: knowledge management is an explicit requirement of the pharmaceutical quality system under ICH Q10, and regulators expect manufacturers to manage product and process knowledge systematically from development through commercial life. When that knowledge is instead scattered across a dozen disconnected repositories, the result is slower issue resolution, harder onboarding, duplicated work, and genuine compliance risk.

Current Situation

Company AC’s real base of product knowledge lacked a shared taxonomy to unify it. The Company launched a Product Knowledge Management Transformation initiative to replace this environment and engaged Iknow to design the taxonomy that would power the new platform’s search and browse experience.

• Product knowledge was scattered across more than a dozen systems.
• The legacy platform was running out of time. The product KM system’s underlying SharePoint 2013 platform was nearing the end of support from both Company AC and Microsoft and had seen no major upgrades since its pilot. Real-world process adherence was around 50%.
• No single taxonomy comprehensively covered pharmaceutical technology. Existing taxonomies in other systems were outdated, and none reflected the full range of pharmaceutical technology concepts needed for search and filtering.
• Regulatory stakes were high. Knowledge management is a defined requirement of the pharmaceutical quality system under ICH Q10, so a fragmented, hard-to-search knowledge base carried compliance risk, not just inefficiency.

• Create a new, integrated taxonomy to categorize pharmaceutical technology product knowledge, drawing on and augmenting existing taxonomies across all relevant knowledge sources.
• Support a unified search and browse experience with filtered, sortable search results, product summary pages, and term-driven pages.
• Incorporate ontology features, such as alternative labels and cross-topic relationships, to improve content tagging and search precision.
• Prepare the taxonomy for integration with Company AC’s existing ontology management, content tagging, and search platforms.

How Iknow Structured the Work

Iknow structured the engagement in four stages — Discovery, Draft & Verification, Final Taxonomy, and Training & Documentation — moving from data gathering and SME interviews through iterative drafting, classification testing, and production readiness.

Key Activities & Decisions

• Discovery and data collection. Iknow collected existing taxonomies, picklists, glossaries, product and site lists, and business process playbooks.
• Stakeholder interviews and ontology evaluation. Iknow interviewed subject matter experts across drug substance, drug product, device, and analytical functions to confirm requirements and capture real-world search use cases. Iknow also evaluated relevant SciBite and FDA ontologies for possible inclusion, covering therapy areas, drug names, and analytical steps.
• First-draft taxonomy structure. Iknow analyzed the data and interview input to prepare an initial structure organized around facets mirroring pharmaceutical technology’s areas of focus — API and drug product, manufacturing steps, equipment and materials, analytical/statistics/chemometrics, devices, and validation — along with event types, document types, sites, and external organizations.
• SME workshops and iteration. Iknow conducted SME workshops to review and refine the first-draft taxonomy, then prepared subsequent iterations directly in Company AC’s ontology and taxonomy platforms using Semaphore.
• Classification testing. Iknow reviewed sample documents to confirm expected topic coverage, compared expected and actual classification results, and worked with Company AC’s IT department to implement the recommended changes, repeating this cycle until performance was satisfactory.
• Production readiness and training. Iknow supported the re-publication of the finalized taxonomy and rules as production-ready, tested bulk classification behavior in the target repository, and prepared an implementation guide, a governance guide, and other training documentation.

Stakeholders & Collaboration

Iknow worked directly with Company AC’s pharmaceutical technology business team, its search group, and the initiative’s business owner, coordinating through a structured statement of work that defined which activities Company AC and Iknow would lead at each stage.

Reconciling Nine-Plus Overlapping, Inconsistent Taxonomies

Company AC’s knowledge was siloed across at least nine platforms, each with a partial or outdated taxonomy, or none at all. Iknow addressed this by inventorying each source’s platform and taxonomy maturity, then drafting a single structure and building the new taxonomy to integrate and augment those sources rather than replace them wholesale.

Designing a Taxonomy Precise Enough Without Becoming Unusable

A taxonomy for the pharmaceutical technology domain risked being either too generic for precise search or too detailed to manage. Iknow addressed this with explicit design principles: a single preferred label with defined alternative label types, a hierarchical structure that limits polyhierarchy, minimized pre-coordination of terms, and targeted associative relationships — for example, linking a product to its dosage form and manufacturing equipment — to keep the taxonomy precise and maintainable.

Quantitative Outcomes

• Content scale addressed: more than 12 million product technical documents across 10+ authoritative repositories.
• Knowledge sources inventoried: 9+ distinct Company AC platforms and taxonomies reviewed and integrated.
• Facets delivered: an integrated taxonomy spanning products, dosage forms, manufacturing steps, equipment and materials, analytics, sites, business units, external organizations, document types, and sources.
• Delivery timeline: taxonomy discovery through production-ready delivery and training completed within roughly seven months.

Qualitative Outcomes

Company AC’s pharmaceutical technology team adopted a single, coherent taxonomy tailored to its areas of focus, replacing a decade-old, unsupported product KM platform and a set of disconnected, partially maintained taxonomies. The new taxonomy was purpose-built to integrate with Company AC’s existing taxonomy management and search investments, enabling the company to build its next-generation knowledge platform using tools it already owned rather than incurring new licensing costs. Company AC’s business case projected that the broader transformation initiative, of which this taxonomy formed the foundation, would deliver meaningful annual cost avoidance and support faster issue resolution, technical transfer, and deviation response, while strengthening adherence to ICH Q10.

Timeline to Impact

Iknow completed the engagement on the agreed timeline, allowing Company AC to begin deploying the new knowledge platform as outlined in its transformation business case.

Core Skills

• Taxonomy design & development
• Search & content classification strategy
• Regulated-industry knowledge management
• Change management & training

Methods & Frameworks

• Structured SME interviews and requirements gathering
• Faceted taxonomy design with controlled labeling and relationship modeling
• Iterative classification testing and validation
• Phased discovery-to-production delivery methodology

Technologies & Tools

• SciBite CENtree ontology management
• Sinequa enterprise search with Semaphore classification
• SharePoint and Veeva Vault content environments