Skip to content
Talk to us
MenuClose

Reinventing professional services for the agentic era.

Foundry is our internal tool for organising engagements, methods, sources and agent-supported work. It supports our practitioners as they deliver services tailored to each organisation.

Foundry kick-off room showing an agent-drafted meeting record, its source-linked output and Morgan Ellis as the human reviewer. Fictional product demonstration.

A room at work: an agent-drafted meeting record awaiting human review. Fictional product demonstration with synthetic data.

How Foundry supports our practitioners

  • Engagements and product work.

    A shared working environment for engagements, our own product work, people and activity waiting on a practitioner.

  • Methods that organise delivery.

    Stages and rooms give an engagement its structure, with work and decisions organised around the method.

  • Sources and draft outputs.

    Briefs, documents and knowledge sit alongside agent-supported work for practitioners to review.

  • People directing the work.

    Our practitioners review outputs and make the decisions about acceptance, further work and handover.

Specialist roles support the work, from finding evidence to drafting and checking. Our practitioners set the task, direct the agents and review the outputs.

  • SourcesEvidence librarian

    Keeps the list of every source: what it is, who owns it, how it is classified and whether it is current.

  • DocumentsDocument compositor

    Puts together the accepted version of a document, stamped with its version, date, coverage and the gaps still open.

  • AnalysisInterview analyst

    Reads workshop notes, procedures and templates and sets out the themes and steps, with citations.

  • AnalysisUse-case analyst

    Compares candidate use cases side by side and drafts the prioritisation from the evidence.

  • EvidenceMaturity evidence mapper

    Maps the evidence onto your capability baseline and shows where it is thin, rather than scoring it.

  • ChecksEvaluation runner

    Runs the agreed test cases on every build and grades each answer: is it right, is it cited, and does it decline when the sources say nothing.

  • BuildBuild engineer

    Writes code and tests against the requirements, for one of our engineers to review and merge.

  • DeliveryMinutes and status drafter

    Drafts the weekly status from the check-ins and the workstream record. The named lead edits and sends it.

Put AI to work in your business.