Set up properly, AI is a toolkit that does sourced research, keeps its cost under control, standardises work and checks its own output. Each programme focuses on one concrete skill, can be taken on its own, and participants leave with a working setup on their own computer.
AI can be convincingly wrong. That is why we train participants not as users but as experts who direct AI and check what it produces.
Participants see live how differently the same task turns out with a poor prompt and with a well-built brief, and start using prompts, plans and multi-step workflows in their own work.
A prompt library and a brief skill
Participants assess the risk of a piece of content before giving it to AI, anonymise it where needed and work safely with open data.
A data check skill
Participants catch the false report and fake source that AI can produce, and connect Firecrawl to carry out current, sourced research.
An installed Firecrawl connection and a research skill
Participants prepare an important decision not in a single chat but with a workflow of agents that research, challenge and review, while keeping the cost under control.
A research and review workflow and an installed Context Mode
Participants prepare a corporate report and an interactive dashboard with prompts, without needing technical tools such as Power BI.
A dashboard built on open data and a report skill
Participants turn a topic into an effective mini presentation in their corporate style, quickly and with prompts.
A mini presentation and a presentation skill
Participants turn a task they repeat constantly (a report, a contract, a proposal) into a skill that AI performs to the same quality every time.
A skill for their own work and a CLAUDE.md file
Participants query, through a single API, how their brand and competitors are perceived by different AI models such as ChatGPT, Gemini and Perplexity, and report the results with Claude.
A brand and competitor visibility report and a monthly tracking skill
Participants turn scattered knowledge into a linked system that lives on their own computer and grows together with AI.
An installed Obsidian notes space and knowledge system
We build the programme around your organisation’s AI Index results. It starts with AI 101 training for a shared language and safe use across the organisation, then we plan department programmes around the weaker dimensions and roles.
Participants learn to describe a task in a short paragraph. The example below runs on fictional data.
Columns: region, month, sales. Write a single-file HTML page that reads the file and shows each region’s total as a bar chart. Send the data nowhere.
For a team not on the list, such as legal, procurement, production planning or R&D, we design the programme from scratch. We draw sample tasks from the team’s real work, prepare exercises with open data and non-sensitive content and deliver them with the tool your organisation allows.