AI training for project managers: status updates people can act on

What the sponsor needs to see first

Suppose an implementation team has finished configuration. Customer testing has not started because the customer has not supplied test accounts. The project manager needs to ask the customer's project lead to arrange those accounts. The launch date remains a target in the plan; it has not been reconfirmed.

A training participant receives those facts and uses AI to prepare a short update. The first draft opens with the completed configuration work and mentions the missing accounts near the end. Every fact may be correct, but the reader has to finish the whole passage to understand why testing has stalled.

The instructor suggests a specific edit: move the missing accounts into the opening sentence and connect that information to a request for the right person. The author revises the draft. A sample update from this exercise could read:

Customer testing has not started because the test accounts are still missing. We need the customer's project lead to arrange those accounts. Configuration is complete. The launch date remains the target in the project plan and has not been reconfirmed.

The sponsor can see both the progress and the action needed. The customer's project lead knows what is being requested. The update does not need an invented risk rating or a new date to be useful. It explains what is known and that accounts are needed before testing can begin.

Write for two readers of the same project

After preparing the short update, the participant writes a version for the delivery team. More detail can help colleagues return to the work: a link to the confirmation that configuration is complete and another to the discussion about test accounts. The exercise materials can supply those records. The team can then check the basis of the update and find the supporting information.

The sponsor's version can stay as brief as the example. There is no need to copy the entire task history into it. When shortening the text, however, the author keeps the reason testing has not started and the qualification about the launch target. Removing those sentences would produce a shorter update that tells the reader less about what the project needs.

A colleague reads one version and explains the next action they take from it. If the answer is simply “we're waiting on the customer,” the author still needs to clarify what is missing and who should arrange it. The instructor might also suggest cutting the configuration description if it dominates the text. The effect of choosing and ordering information becomes visible in the reader's response.

This is a useful point to compare the author's version with the AI draft. Shorter sentences or a clearer explanation of cause and effect may be worth keeping. The project manager selects the edits that help the intended reader understand the situation. Any judgment about whether the launch date is achievable rests on project information.

The next update needs to reflect what changed

In the next part of the exercise, the accounts have been supplied and customer testing has begun. A new issue has been reported, but its effect on launch has not yet been assessed. The participant removes the request for accounts, records the start of testing, and describes the open issue using the information provided. Adding a sentence beneath last week's text is insufficient: the reader would still see an outdated request.

The instructor and author compare the two versions. Can the reader tell what changed? Is it clear that supplying the accounts resolved one problem while the effect of the new issue remains unknown? That discussion helps the PM write successive updates as a continuing account of the project. It also provides practice removing sentences that were useful a week earlier.

Practice with the tools and material the team uses

A new system is not necessarily required. Asana's Smart status documentation describes AI-generated project updates, a choice of reporting period, and guidance for the draft. Users can edit the text before posting it. Those capabilities provide a starting point for discussing tools the company already has.

In the test-account example, check whether the customer's information is available when the update is drafted. A drafting feature does not by itself establish access to separate correspondence. If the information sits only in the PM's email, it needs to be considered when preparing the text, in accordance with company rules for using the tool. Fictional records can support the exercise; using real project material requires approval for the chosen application.

Someone new to AI can practice preparing and revising a first update in that tool. An experienced PM can work from messier notes and ask a peer what action they infer from the resulting passage. A PMO representative or sponsor can discuss the purpose of the weekly update with the instructor. Feedback then reflects the reader's needs.

Sessions can fit around the normal reporting cycle, leaving time for each person to write and discuss their own update. The number of meetings and opportunities to revise work after feedback need to be agreed in the program. Sometimes simplifying an unclear template is enough. If current project decisions are never recorded, writing practice cannot supply the missing information.

Start the training conversation with one update

Syntalith can tailor team training to the updates your project managers write. A proposed workshop includes drafting with AI and revising after a conversation with the intended reader. Participants can also prepare two versions of the same update: a short one for the sponsor and a more detailed one for the delivery team. Exercises and feedback arrangements are agreed before the sessions.

We also build applications and automations. If gathering information from several systems takes most of the effort, connecting those systems is worth a separate discussion. Our article on automating professional services project closure describes that kind of work at a different stage of a project.

To begin, describe an update that left the sponsor asking what was actually happening. Explain which question went unanswered and where the PM looked for the information. Service information is available on Syntalith's pricing page.