01
The AI landscape
Which tool, for which job.
Chat, agents, image, video, voice and automation tools, compared on real tasks so your team picks with confidence.
- Compared on your tasks
- Free versus paid, honestly
- Updated as tools ship
05 / 05 · 8 projects
Training & setup
Hands-on training and setup across the whole AI stack. If it ships, we teach it.
02
Team programmes
From curious to capable.
Short, hands-on sessions and videos: chat or agent, choosing a model, prompting, automation and safe use.
- Hands-on, on your work
- Short videos to revisit
- A playbook to keep
03
Agent workspace setup
Ready on day one.
Claude Code and Cursor workspaces set up with your rules, skills and connectors, so agents work the way you do.
- Rules and skills installed
- Connectors, least privilege
- Your folder conventions
04
Knowledge system
One source of truth for every AI.
Facts tagged verified or assumed, procedures and rules that every AI tool reads, with private data kept private.
- Verified versus assumed
- Procedures as files
- Private stays private
05
Custom skills
Your best work, on call.
Skills your team calls by name: proposals, reports, brand voice, 3D viewers, each in your own format.
- Your formats, built in
- Called by name
- Shared across the team
06
Connectors and permissions
Plugged in, on your terms.
AI connected to mail, calendar, files and CRM, with a switch and a scope for each one.
- Read or write, per tool
- Sensitive tools off by default
- Every action logged
07
Prompt playbooks
Better questions, better answers.
Before-and-after prompts for your real tasks, so the whole team gets usable answers on the first try.
- Built from your tasks
- Before and after
- Checks built in
08
Safe AI by default
Useful, and under control.
Hidden instructions blocked, least privilege for every tool, client data kept out of public tools and a human on every send.
- Prompt injection defences
- Least privilege
- Human approval on sends