What we do

Three offers, one system.

Workflow automation, training, and research and evaluation, from cognitive scientists who build what they recommend. Take any one of them on its own; combined around what your organisation knows and how its people judge, they work as one.

Your organisation, at the centre.

At the centre are your organisation’s positions, evidence, past work and voice, and the judgement of the people who use them. Each offer strengthens one part of that.

A diagram of the three offers working as one system. At the centre is your organisation’s knowledge: your documents, past work and know-how, in plain files. Workflow automation draws on your facts and tone of voice to draft the routine work and read the news, Parliament and funders, and sends its drafts to your people, who decide what goes out. Training builds your people’s capability to think with AI, and teaches them how to use AI well, and judge its work. Research and evaluation designs and measures any program you run, and what your training or services changed goes back into your organisation’s knowledge. What your people change in each draft is kept for next time, so tomorrow’s drafts know about today.

What your people change in a draft is saved, so tomorrow’s drafts know about today.
Does keeping it current mean more work for my team?

It stays current because people use it. A small handful of automations do the filing: a wrap-up at the end of each day, an inbox that files new material, a briefing before each meeting. Records update as people work, without anyone filling in a form.

More on how it stays current

How the parts feed each other.

Each part makes the others better. What your organisation knows feeds every draft. People trained to judge well review them. Evaluation shows what changed. And what your people change in a draft is saved, so tomorrow’s drafts know about today. Bought as separate tools, each would need someone to keep it fed; built as one, the everyday work keeps it current.

Training is what keeps people using a system well, and evaluation is how you know what any of it changed. When evaluation is part of the work, we design it in from the start.

Why put your organisation’s knowledge at the centre?

AI only does good work with the right context, and most organisations supply it by hand, every time. Put it at the centre, and every part of the system starts from it.

How we build it

The same science in each.

Each offer puts the science of how people think, learn and decide to work in a different place. Each has one part we always keep, because it’s what makes the offer work.

Workflow automation

Takes the routine off your desks. Submissions, reports, newsletters, board papers, briefings: the work that comes round every week, drafted from your organisation’s knowledge.

The part we always keep: Nothing goes out until a named person approves it, and each draft prompts them on what to check. The software won’t let anything skip that step.

Training

Builds your people’s capability: to see where AI helps in their own work, to think with it, and to judge what it gives them. For whole teams, one to one, or in a single talk.

The part we always keep: Learning spread over weeks, with practice recalling it from memory, because that’s what makes it last. We never cut it to save a session.

Research and evaluation

Shows what changed. We design and measure any program you run, AI or otherwise, build surveys and measures your team can run itself, and bring the evidence on what makes a submission persuade. More on that

The part we always keep: A comparison wherever one is possible, so you can tell what the program changed from what would have happened anyway.

Where to start.

Start where the pressure is. Training and research and evaluation can each be bought on their own. A workflow automation build always includes training for the people who use it. Whichever you start with, the others connect when you add them.

A good first step is one piece of work that comes round every week, or a training series for your team. Book a conversation

The same three answer a larger organisation’s problems too: the AI licences are in but the gains have stalled, or nobody can say what last year’s training changed. Organisations and their teams

We don’t promise numbers.

Time back for the work that needs judgement is what we work towards, not a figure we guarantee. When we measure, we report what the evidence carries.

Start with a conversation.

Bring the work that takes most of your week, and we’ll talk through where to start.