Putting it together · for leaders
Building AI workflows as a leader
Much leadership work repeats on a schedule. Board updates arrive each quarter. Performance reviews run twice a year. Weekly leadership meetings need preparation. Starting each of these from a blank prompt wastes time and produces uneven results.
Putting it together means breaking recurring work into stages. Each stage gets its own instructions, and a person reviews the work at fixed points. For leaders, it also means designing the process the team follows, so that AI use is consistent and checked across the group.
Where it comes up
A quarterly board reporting cycle
Collect approved figures from finance, draft each section from a saved template, check every figure against the source, and write the forward-looking section yourself. Each step has an owner.
Preparing for a weekly leadership meeting
Summarize each team’s written update in a fixed format, list open decisions, and flag items that need discussion. You choose the agenda from that list.
A performance review cycle for your team
Gather each person’s goals, peer input, and your own notes. Use AI to organize evidence against the goals, not to judge. Write the assessment yourself, then check it for consistency across the team.
What it looks like at each band
E1 · Foundational
Uses AI for single tasks and connects the steps by hand. Each board update or review starts from scratch. Quality varies from one cycle to the next.
E2 · Proficient
Runs recurring work as a saved sequence of steps. Reuses templates and requests that worked before. Reviews the output at a set point before it moves on.
E3 · Distinguished
Designs workflows with clear handoffs between AI steps and human judgment. Assigns owners to each step on the team. Documents the process so a deputy could run it in their absence.
E4 · Exceptional
Builds the leadership team’s operating processes with AI in them, and reviews how they perform: where errors enter, which steps save time, and which need more checking. Removes steps that no longer help. Other leaders adopt these processes.
Common mistakes
One request for the whole board pack
Asking for an entire board update in one request gives the model too many jobs. Draft section by section, against approved inputs.
No review point before distribution
A workflow that sends documents to directors or staff without a person reading them will eventually send an error to the people who matter most.
Letting the workflow make people decisions
A process can gather and organize evidence about performance. The judgment about a person must stay with a manager who knows them.
A worked example
Producing the quarterly investor update each quarter.
Each quarter, paste last quarter’s update and some new numbers into a chat and ask for this quarter’s version.
Old figures and phrasing carry over without notice. The format drifts. Nobody can tell which figures were checked, and the process depends on one person’s chat session.
Step 1: finance supplies approved figures in a fixed sheet. Step 2: a saved request drafts each section from that sheet and the quarter’s key events. Step 3: you write the outlook and the asks yourself. Step 4: finance checks every figure in the draft against the sheet. Step 5: you read the whole update as an investor would, then send.
Each step has one job and one owner. The judgment about outlook and asks stays with the leader. Figures are checked by the team that owns them, and the process can run again next quarter.
The full study guide for this skill, for every role, is Putting it together. The other five skills for leaders are on the AI skills for leaders page. A one-page summary is in the AI workflow checklist.
See where you stand
The free check grades putting it together and the other five skills in about fifteen minutes. Nobody sees the grade but you.
Then build the skills and prove it. The three steps for leaders list the courses and templates that fit the work.