An AI-augmented professional capability company

Turn scattered AI use into a workflow you can repeat, check and defend.

Most professionals now have AI access. Far fewer have a workflow they can operate repeatedly, evidence they can put in front of a client or a board, or capability they can actually demonstrate. Edney Learn helps you build reusable systems around your real work, then measure where you are, build the evidence, and keep the capability.

The AI Capability Benchmark is a developmental profile: a Controlled Developmental Release, not a validated test. It opens November 2026; you can register interest now. It is one way in, not a required first step.

Current focus vertical Pilot

AI-Augmented Project Manager

Project managers were among the first professionals handed AI tools, and among the first to discover the limit. AI drafts a status report, summarises a risk log, rewrites a charter. What it does not do on its own is give you an integrated project workflow, a governance position you can defend, or evidence that any of it improved delivery.

We are running this as a pilot: a small cohort, working on live projects, building a project workflow they own and can keep operating after the programme ends. You leave with one documented project workflow — defined AI entry points, verification steps, decision ownership and stop conditions — and a governance position you can put in front of a steering group.

What "pilot" means here

  • Cohort places are limited and by application.
  • The workflow you build is yours. It is not our product, and it does not become our IP.
  • We are still gathering field evidence. We make no claims about outcomes we have not measured.

The problem

Access to AI is not the same as capability with AI.

Three years of experimentation has produced a great deal of activity and remarkably little that can be measured, repeated, or shown to anyone.

01

The output is not the asset

A good AI-assisted document is worth something once. The workflow that produced it is worth something every time. Most professionals have accumulated outputs and no system.

02

Quality is assumed, not engineered

Without explicit standards, verification steps and defined review gates, quality depends on whether the person noticed the problem that day. That is not a professional control.

03

Nothing can be demonstrated

Ask a professional to evidence their AI capability and most can show artefacts, not capability. There is no record of judgment applied, contribution allocated, or output verified.

The distinction we work from

AI-enabled is access to tools and assistance with isolated tasks. AI-native is the operating design we build: a human and AI working through one governed workflow with explicit authority, handoffs, verification and stop conditions. AI-augmented is the outcome: a professional who can responsibly perform an expanded role because that workflow is operable. AI-native is the mechanism; AI-augmented is the result. Read the full distinction →

Why our workflows are different

Generic AI, used generically, produces generic work.

The differentiator was never access to the model. Everyone has that. The differentiator is the codified domain expertise fed into the workflow: the standards, the sequence, the checks, the things a practitioner knows to look for and a general-purpose prompt does not.

Every Edney Learn workflow product carries codified best-practice from the relevant professional domain. That is how the product is built.

How the result actually happens

We bring the method and the professional controls. You bring the context, judgment and live work. The result depends on both, which is why we describe codified best-practice as how the product is built, not as a promise about what you will produce.

The method

Design the workflow once. Operate it many times.

Our ten-stage AI-Augmented Workflow Design Method is a method for designing or redesigning a workflow. It is not a checklist you run from Stage 1 to Stage 10 for every task you do.

Design

The ten-stage method takes you from target output to a documented operating procedure — once.

Operate

You then run the workflow the method produced, repeatedly, as normal professional work.

Collect evidence

Operation generates a record: judgment applied, AI contribution, validation decisions, output quality.

Optimise

That evidence feeds the next redesign cycle. The workflow improves because it is measured.

Escalation is a control, not a stage

Every workflow we help design carries explicit escalation rules — designed within Engineer and owned within Govern. When source material is insufficient, when a decision requires accountable human judgment, when rights or confidentiality are in question, when a claim would outrun its evidence, or when the matter is legal, medical, financial or otherwise regulated — the workflow stops and returns to a named person. It does not proceed and hope.

See the ten-stage method

Escalation, in practice

A workflow that cannot stop itself is not a professional workflow. Designing the stop conditions is as much of the work as designing the steps, and it is the part that generic prompting has no concept of.

Role pathways

Choose the workflow closest to your work.

Same method and standards, configured around the outputs, decisions and quality bars in your profession. Start with the route closest to your work. Each one carries an honest status label.

Available and running now

Not ready yet, but we can work with you on a scoped basis

Available Enquire and start today
Pilot Running now, by application
In Development Register interest
Future Signal demand only

The shared offer architecture

Start with the support you need.

Enter at the diagnostic, a workflow pack, a workshop or bounded coaching — whichever fits — and stop where the value stops. Nothing obliges you to buy the next stage, and nothing advances you automatically to a licence or credential.

See programmes and workflow packs

Available now

  1. 01

    Developmental diagnostic

    Baseline your capability for reflection and pathway fit.

    Pilot / CDR
  2. 02

    DIY course or workflow pack

    Self-study implementation of a defined role-specific workflow.

    Available starting 1 October 2026

    Available
  3. 03

    Live implementation workshop

    Build it with us, using your own live professional work.

    Available
  4. 04

    Coaching and review

    Bounded support for hard decisions and quality review.

    Available

Future architecture, not for sale today

  1. 05

    Licence

    Institutional deployment for organisations and educators.

    Future
  2. 06

    Microcredential and evidence portfolio

    Assessment of demonstrated capability — future architecture.

    Future

Evidence discipline

We separate what we can prove from what we have merely done.

  • The method and the product designs exist, are documented, and we operate them ourselves.
  • We do not yet publish customer outcome evidence: no testimonials, no case studies, no performance figures.
  • We will update our claims only when we hold consented, documented customer evidence. Until then, inventing them is the one thing we will not do.

Read the full evidence policy →

Founder and company

Built by a practitioner, then a researcher.

Edney Learn was founded by Dr Jim Choo. Before the doctorate came two decades in industry: electronics business development, business analysis and process re-engineering, IT project management including ERP deployment, enterprise data governance and R&D programme management, at Compaq, Agilent Technologies and Keysight Technologies.

The method comes out of that practitioner experience, then tested against a PhD in Curriculum and Pedagogy Innovation on how professional capability actually develops. That is the ground the workflow method, the diagnostic and the evidence discipline are built on.

The method comes out of twenty years of delivery work, where a specification either holds or it does not, then tested against research on how capability actually develops.

Start here

Start by measuring where you actually are.

When open, the AI Capability Benchmark will take a baseline across task-selection judgment, human–AI allocation, verification practice, workflow integration, accountability and evidence, then suggest which pathway fits. It is one way in, not a required first step.