Lindy puts an AI employee inside Slack and declares the agent era over
The four-year-old startup is betting that enterprises want to hire an AI coworker, not configure an agent platform. The packaging shift matters as much as the product.
Lindy launched Lindy Teammate on Sunday, an AI coworker that lives inside Slack, joins meetings, and takes work from anyone on the team. Founder Flo Crivello framed the release in blunt terms: the company is killing the AI agent and introducing the AI employee.
The mechanics, as Lindy describes them, are deliberately unremarkable. You add Teammate to Slack channels the way you would add a new hire. Ask in any channel and it responds in thread. It attends video calls on Meet, Zoom, and Teams, then records, transcribes, and files the meetings in folders the team can query later. It connects to company tools and carries a persistent memory of the business, which the company's site says lives in plain, editable files users can read and correct. Crivello claims teams get 10x more done with it, a figure that is his and not anyone's audit.
Packaging is the product
Strip the launch language and Teammate is an orchestration layer with Slack as its front door. That is precisely the point. Lindy spent four years building an agent platform with a builder, integrations, and workflow tooling, and has now concluded that enterprises do not want to configure any of it. They want to message a coworker. The company positions this as the biggest launch in its history, and the framing puts it in direct competition with the digital-workforce push we covered at ServiceNow, where the sales motion is also headcount-shaped rather than platform-shaped.
The economics underneath remain metered. Lindy's published tiers start at $29.99 per month for 3,000 credits and run to $199.99 for 35,000, with an enterprise tier adding HIPAA support with a signed BAA, SOC 2, and audit logs. Selling an employee while billing in credits is the tension every AI employee vendor now carries. Buyers benchmarking against salaries will eventually push vendors toward outcome pricing of the kind HubSpot adopted for support resolutions, and a credit meter sits awkwardly next to the hiring metaphor.
What buyers should verify
Everything above comes from Lindy itself. There is no independent benchmark of Teammate's reliability, no published customer case study with numbers, and no third-party review of how its memory behaves when two departments tell it conflicting things. The company's own guide to AI employees is candid that these systems suit routine work and should not replace human judgment in ambiguous situations, which is a more honest boundary than the launch copy draws.
Still, the direction is clear and the timing is not random. Slack is where enterprise work already happens, and an agent that requires no setup, no workflow design, and no training beyond an invite removes the deployment friction that has stalled agent rollouts all year. If Teammate works even close to the way its demo suggests, the interesting question for the GaaS market is not whether the AI employee framing sticks. It is how fast every other agent vendor repackages what they already sell to match it, and what happens to per-credit pricing when the thing being priced is supposed to be a colleague.