Skip to main content
Foundation
Capabilities

The technical capabilities behind the work.

A small set of distinct technical capabilities underpins everything we deliver. Each is applied across more than one part of the work, and the engagements compose them in different ways.

  • 01

    Ingestion & standardisation

    We bring workforce data together from across the systems an organisation runs and resolve it onto a single schema, reconciling inconsistent fields and formats and collapsing the same role recorded many times into one record. Where structured data is thin, the same layer takes operating documents and interviews as an alternate input.

    Used in. Data standardisation, Role deduplication, Process mapping

  • 02

    Inference over documents

    We read job-related documents and extract the units of work they describe, both those stated explicitly and those implied, then reconcile against our reference datasets so the output stays aligned to a known market vocabulary. The same primitive runs over skills and over tasks.

    Used in. Skill inference, Task inference

  • 03

    Enrichment & decomposition

    We add the structured attributes each inferred unit needs to be reasoned over, and for tasks we decompose the work into discrete actions, the granularity at which AI impact is measured.

    Used in. Skill enrichment, Task enrichment

  • 04

    Hierarchy construction

    We take a large, flat set of data points and build a governed hierarchy over it, grouping by shared structure, not by label. The output is a standalone asset that holds across thousands of items.

    Used in. Skill taxonomy, Task taxonomy, Job families

  • 05

    Scoring against a rubric

    We score every item in a population against a defined rubric and return the reasoning alongside each score, so results stay consistent across thousands of items and stand up to review. The rubric can be one of ours or the client's own.

    Used in. Role levelling & evaluation, Automation assessment

  • 06

    Generation from patterns

    We turn the strongest scored opportunities into structured artefacts built around our best-practice working loops, each tied back to the specific work it changes.

    Used in. AI use-case inventory

  • 07

    Retrieval & matching

    We match each opportunity against our maintained catalogue of AI products, scoring fit against the client's context to surface the closest options with their reasoning. The catalogue is curated and kept current, so matches reflect the live market.

    Used in. Vendor & product recommendations

  • 08

    Agent engineering

    We take a matched opportunity from a high-level design through to a build-ready specification, then engineer, integrate and deploy the agent against the real workflow and the systems it depends on. The same capability spans the full path, from shortlist design to a working system in production.

    Used in. Agent design, Agent build & deployment

See how these capabilities come together in an engagement.

Back to Foundation