Truemed, a 50-person health payments company, had its head of support build Lindy agents that classify and answer routine tickets and route the rest, without engineering help.
- First reported
- lindy.ai · 1 Oct 2025
- Added
- 13 Oct 2026
- Source checked
- 13 Oct 2026
What they did
Engineering had no capacity, so the head of support built the agents on their own. They classify incoming inquiries, answer routine ones such as payment and eligibility questions, and route or escalate the others. Replies are kept consistent with health payment rules in a tightly regulated market, and the same approach was used for lead management and onboarding workflows.
What happened
The company says about 36% of support volume is now handled without a human and the cost of each automated ticket fell by about two-thirds, with no extra support hires.
Keep in mind
Figures are company-reported on a vendor page with no time period or method, and the cost cut applies only to automated tickets. The page does not name the channels or helpdesk used.
Try it with your agent
Watch new tickets in <your helpdesk or support inbox>. Classify each one into <your list of ticket types>. For types marked routine in <your FAQ and policy document>, draft a reply that follows the document word for word on rules and eligibility. Route everything else to <the right person or queue> with a one-line summary. Send only the routine replies I have approved as a category; hold anything about refunds or compliance for a human.
Replace the <angle brackets> with your own details.
Original source
lindy.aiVendor case studyPublished by the company that makes the agent. Read it as a claim.
Tools for this kind of work
The whole toolkit →
