AI features in Projektiassistent: a video guide

The most common assumption about AI in project management is that it writes text. In Projektiassistent that is one of five things the AI does — and the least interesting one. The video guide above plays through all five on real screens, using a single example project.

The video runs a little over two minutes and is split into chapters, so you can jump straight to the feature you care about. The same material is written out below.

1. Attention signals

This block does not wait to be asked. The AI reads the project data — task deadlines, risk scores, workflow SLAs, the critical path — and surfaces the places that need a decision: overdue tasks, active high risks, blockers. Every row carries a priority, so you do not have to work out where to start.

The difference from an ordinary task list is that only things computed from the project land here. If everything is fine, the block is empty — and that is an answer too.

2. A status summary for management

Writing the status report is where a project manager spends the most time for the least value: the data already exists, it just has to be moved into another shape. The AI assembles the summary from the data in the project — what is done, where the budget stands, which risks are open — and you can edit the text freely before sending it.

The summary is included in every plan. The document exports to PDF or DOCX with five design templates, so it fits into a board pack without reformatting.

3. A grant application from the same plan

This is the feature most users actually come for. The AI takes the existing project plan and drafts a funding application from it: the applicant, the problem and the need, the activity plan and schedule, the expected impact. The video shows the finished text section by section.

Two honest limits. First, the funding body decides on eligibility — the draft has to be reviewed before submission; this is not an automatic “submit and hope”. Second, the feature is not included in the trial; Standard gives three applications a month and Business is unlimited. More on the grant writing page.

4. The critical path

The critical path is the longest chain of linked activities, and it sets the shortest possible duration of the project. If a task on that chain slips by a day, the end of the whole project slips by a day. The system works the chain out from the tasks and dependencies and shows how many tasks are on it and what the projected end date is.

The practical value is simple: when you have to choose where to put your attention, the answer is here — the other tasks have slack, the ones on the critical path do not.

5. The AI assistant and what-if simulation

The AI project manager assistant gathers project health, the critical path and the latest simulations into one management view. AI recommendations point out missing activities, signs of overload and risk areas.

What-if simulation answers the question people actually ask in a meeting: “what happens if this task is ten days late?” You enter the delay and see the effect on the project end date — before you promise the client a new one.

Meeting notes → actions takes the text of your notes and finds the tasks, decisions and risks in it, ready to add to the project in one click. These three are part of the Business plan.

What the AI does not do

The honest list is short but important. The AI does not make your decisions — it prepares them. It can be wrong, like any AI, so the plan and the application have to be read through. And it knows nothing you did not give it: the more precise your input and the more documents you attach, the more usable the result.

Where to go next

The written chapter with every field and button is in the user guide: AI features. The product view is on the AI project manager page, and if you would rather see how a plan comes about in the first place, start with the project plan video guide.