This CEO has AI helpers run his inbox, calendar and rival watching

Asana CEO Dan Rogers, 49, lets AI agents help run his workday. They sort his email, check his calendar, report on sales and watch rival companies. Asana sells work management software, so he is using what his company makes. His main warning is that the agents take effort before they pay off.

What the agents do

Every morning, Rogers asks three agents he calls his personal productivity teammates to plan his day. One scans his Gmail for messages that need action now. Another checks his calendar and alerts him if his time is not going to his priorities. The third reads meeting transcripts to find new commitments he made.

• A customer agent flags issues, such as a business that delayed renewing its subscription.
• A sales questions agent shows what employees struggle to understand, like a new pricing model.
• A performance agent turns internal data into reports on sales and new customer sign-ups.
• An engineering agent suggests how to use engineers' spare time and tracks requested features.

The rival watcher

One agent tracks competing software companies. It follows their product announcements and what people say about them on social media, then writes daily briefings. Rogers reads them every three or four days. It also talks to an agent used by Asana's chief product officer. If a rival launches a feature, the two compare it with Asana's current or planned products and flag anything that might change the product plan.

What went wrong at first

The rival agent started out annoying. It flagged information Rogers did not care about. It became useful only after he gave it more precise instructions. "Some of these agents do require work to get them where you want them to be," he told Business Insider.

Keeping it safe

Some agents handle sensitive company data. Rogers said Asana limits what data they can reach and what actions they can take without a human approving it. "Security is paramount," he said. He also said work agents differ from home ones, such as those that book hotels, because they are built to understand a company's knowledge and handle private information.

What to do

Anyone curious about AI helpers at work can follow the same pattern. Start with one narrow job. Write clear instructions and refine them when the results miss. Limit what the agent can see and what it can do on its own.