Scaling Linear's GTM to $100M+ ARR — Cristina Cordova, COO
- Source
- youtube.com
- Author
- Latent Space
- Date
Context
More than half the issues created in Linear, the tool many software teams use to file, assign and track work, now come from AI agents rather than people, according to its chief operating officer, Cristina Cordova, on the Latent Space podcast (episode page with transcript). She dates the shift from Linear's launch of an server in 2025, a connector that lets outside agents read and write Linear data directly, and says the change took about a year and a half. The episode's notes put agents' share a year earlier at 3%.
That raises a question for any app whose data agents can reach: if agents do the writing, why would people open the app? Cordova argues that being a database behind an MCP server, handing data to outside agents, will not make a good business. In her view a product has to bring workflows inside and make them better than the outside tools that connect to it, or people have no reason to come in. Linear watches two lines, MCP usage and app usage, and she says both are rising, because a chat window is a poor way to see a whole team's work and is usually used by one person at a time.
The argument connects to a March post by Linear's chief executive, Karri Saarinen, titled Issue tracking is dead. Cordova reads it narrowly: the issue remains Linear's basic unit of work, and what is dead is people moving each bug by hand through reporting, labeling, assigning and closing. Linear now describes itself as a product development system in which agents do those steps, including a feature called Loops that can assign an agent to an issue and have it start coding.
The evidence for her test is her own account. She gave no app-usage figures in the conversation, and when the interviewer asked whether rising app usage simply reflected a growing customer base, she said Linear looks at several factors, mainly app sessions, without giving numbers. Her claim that Linear is escaping the database fate she warns about therefore rests on figures Linear has not published.
- MCP — The Model Context Protocol — an open standard that lets any AI assistant plug into any tool or data source without custom integration code.
- AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.
videoHow do you diffuse AI into the real world? — Varun Shenoy, Long LakeAI Engineer
podcastAnthropic’s first technical PM on token maxing, the jagged edge, and living in the future | Dianne PennLenny Rachitsky
articleHow Anthropic ships the #1 LLM in the World safely: Separating the Brain from the HandsLatent.Space
Checking sign-in…
Loading comments…



