Agents we run ourselves
Before we build agents for anyone else, we run them in our own operations. These are in regular use today.
Receipt-filing agent
Reads each uploaded receipt or invoice, names it, labels the expense and files it into the company's folders.
Two-model manuscript screening
Two AI reviewers on different models assess each submission against written criteria. An editor agent reconciles their reports for a person to decide.
Submission review agent
Reviews article submissions against a written editorial standard, documents the changes needed and prepares the email to the author.
Comment moderation agent
Screens reader comments before they appear and holds borderline ones for a person.
Voice tutor
Asks questions aloud, listens to the spoken answer and checks it against a reference answer.
Publishing agents
Research, draft and check articles on a schedule. Nothing goes live until it passes written sourcing and quality checks.
Reporting agent
Pulls site and search data each week and emails a report that separates engaged visitors from bot traffic.