Four capabilities. One lifecycle.

End-to-end content engineering across information architecture, migration, CCMS implementation, and AI-ready retrieval. Most engagements move through more than one pillar.

The matrix

The shape of an engagement.

Each pillar concentrates work in different methodology phases. The shape of an Extense engagement is how those distributions overlap.

How each capability pillar concentrates its work across the five methodology phases. Cells marked as primary-load (visually shaded and noted in copy) indicate where the pillar's primary execution lives.
Pillar × Phase Discovery Audit, stakeholder interviews, current-state mapping Architecture Schema, taxonomy, structural decisions Migration Conversion, normalization, content recovery Implementation Configuration, build pipelines, rollout Enablement Training, governance, runbook handover
Information Architecture Schema and taxonomy that determine downstream possibility Stakeholder interviews, content audit, use-case mapping. Primary-load phase for this pillar. Content models (DITA elements, S1000D modules, custom DTDs), taxonomy, reuse strategy. Primary-load phase for this pillar. IA validation through pilot conversion of representative content. IA propagation across authoring tools and CCMS configuration. Governance handover, IA documentation, author orientation.
Content Migration Recovery operation that turns legacy content into reusable assets Content audit, inventory, source-format assessment. Conversion design, target schema validation, QA harness design. Primary-load phase for this pillar. Pilot conversion, production batches, normalization, metadata enrichment. Primary-load phase for this pillar. Cutover, parallel running, rollback procedures. Author transition, runbook handover, post-migration optimization.
CCMS & Publishing Operating system for content plus the CI/CD that closes the loop Current-state assessment, platform fit analysis, requirements gathering. CCMS configuration design, workflow design, publishing pipeline architecture. Primary-load phase for this pillar. Data load and content migration into the configured CCMS. Pipeline build, CI/CD integration, automation rollout, multi-format publishing. Primary-load phase for this pillar. Admin and author training, runbook handover, ongoing optimization. Primary-load phase for this pillar.
AI-Ready Content Translation layer between structured content and AI retrieval Content readiness assessment, retrieval use-case mapping. Chunking strategy, metadata schema, retrieval architecture. Primary-load phase for this pillar. Content remediation for AI ingestion (chunking and metadata enrichment). Pipeline integration, retrieval evaluation harness, RAG deployment. Primary-load phase for this pillar. Ongoing tuning processes, AI-ready authoring practices. Primary-load phase for this pillar.

How an engagement moves through the lifecycle.

IA work concentrates early, but it never fully closes. The decisions made in Architecture — schema design, taxonomy, reuse strategy — surface as constraints during Migration, configurations during Implementation, and retrieval boundaries during AI enablement. When IA is right, downstream phases inherit a foundation that holds. When it isn't, the same work has to be re-litigated in every phase.

Migration work depends on Architecture decisions made in IA. The conversion engine is only as good as the target schema it's converting toward — and most failed migrations failed because the target was unspecified or wrong before the first batch ran. The Migration phase carries the heaviest production load, but Architecture carries serious load too.

CCMS Implementation requires Architecture decisions made during Migration. The content model that lands in the CCMS shapes workflows, permissions, publishing pipelines. Get the IA wrong and the CCMS configuration fights the authors. Get the Migration wrong and the CCMS launches with the same authoring problems the legacy system had.

AI-Ready Architecture often runs in parallel with CCMS Architecture, not after. Chunking strategy, metadata schema, and retrieval architecture are content-engineering decisions that interact with the CCMS configuration; they shouldn't be retrofitted. The most successful AI deployments are the ones where the content was engineered for retrieval before the model was selected.

Enablement runs through every phase, not at the end. Discovery teaches the practitioner how the program works today. Architecture documentation becomes the team's runbook. Migration training transitions authors. CCMS rollout becomes the operating manual. AI-Ready evaluation becomes ongoing tuning. The handover isn't a phase — it's a discipline.

An end-to-end engagement.

A Class III medical device manufacturer consolidating regulatory documentation across global markets, migrating from FrameMaker to a DITA-based CCMS with RAG enablement on top.

The situation

Decades of FrameMaker IFUs across multiple device families, each with regulatory variants for FDA, EU, Australia, and dozens of additional markets. Translation costs trending upward year-over-year. A new AI-assisted compliance assistant initiative running into recall-risk concerns because the content estate couldn't be cleanly retrieved. Documentation team running parallel workflows that weren't sustainable.

The work

IA design first: device-family content model with regulatory-variant profiling, controlled vocabulary aligned to MedDRA, reuse strategy across IFUs and labels. Migration phase converted legacy FrameMaker estate to clean DITA at scale, with a QA harness that surfaced pre-existing content errors as part of the migration. CCMS implementation deployed configured workflows for regulated authoring. AI-Ready phase architected chunking and metadata for the compliance assistant pipeline, with retrieval evaluation against representative queries before model deployment.

The outcome

Material translation savings on subsequent regulatory submissions. Faster publishing across markets via CI/CD-driven build pipelines. Compliance assistant deployed against a content estate engineered for retrieval, rather than retrofitted afterward. Documentation team operating a single source rather than parallel workflows.

Anonymized for client confidentiality. Specific scope, contract details, and named outcomes available under appropriate NDA channels.

Go deep on a single pillar.

Sample Content Assessment

Submit a 20-page sample. We'll return conversion feasibility, content recovery rate, and engineering effort within two business days. The analysis is the basis for any engagement scope, single-pillar through full lifecycle.

Submit a sample →