Role Clarity Card
Define why a role exists, the outcomes it owns, the decisions it can make and where it must collaborate.
You leave withA one-page role agreement that is more useful than a long job description.
Use this toolkit Resources
Tools for the work. Knowledge from the field.
Use practical team toolkits today, then follow our AI field notes as we test models, apply them to real workflows and document what is genuinely useful.
Founder toolkits
Practical structures for work your team needs to own.
Every toolkit includes implementation guidance, a worked example and a blank template. Start with one live problem, use the tool with your team and adapt it to how your business actually works.
013 toolkits 022 toolkits 032 toolkits 044 toolkits 051 toolkit
Clarity and accountability
Make ownership, decisions and delegated outcomes visible.
Repeatable workflows
Turn recurring work into a process the team can follow and improve.
Performance rhythms
Track the signals that matter and run reviews that produce decisions.
Manager routines
Give managers practical structures for onboarding, coaching and learning.
Responsible AI
Choose useful AI work and define the human controls around it.
What every toolkit includes
Clear guidance. Know when and how to use it.
Worked example. See the right level of detail.
Usable template. Copy, adapt or print it.
Browse all 12 toolkitsAI field notes and knowledge
Testing models against the work businesses actually do.
This is where we share what we learn while testing different AI models and applying them to real operating work. Published notes will show the task, method, result and limitations so founders can judge what is useful for themselves.
Why systems thinkers are the quiet backbone of every great business.
Before comparing models or automating work, understand the decisions, handoffs and feedback loops that make the work function.
Read the articleModel tests
The same real-work task tested across different AI models, with the inputs, output quality, limitations and useful differences made visible.
Real-work use cases
Examples of where AI helps inside sales, service, delivery, reporting and management workflows, including where human judgement must stay.
Implementation notes
What happens after a promising demo: data boundaries, workflow design, team adoption, review points and evidence of actual value.
How the notes will be written
Specific task. What real work was tested and why.
Visible method. The model, inputs and review standard.
Honest result. What improved, failed or stayed human.
Visit all insightsWhen self-implementation reaches its limit
Some problems sit across roles, systems and decisions.
Use the toolkits and field notes freely. If implementation reveals a deeper operating constraint, contact us for help diagnosing, rebuilding and testing the system with your team.