Flo Crivello

Flo Crivello is the founder and chief executive of Lindy, a company developing AI employees that act across calendars, inboxes, business applications, and the web. His central proposition is that capable software agents can give individuals and small teams operational capabilities previously requiring entire organizations.

Crivello worked in software engineering before moving into product management at Uber, a transition he pursued to prepare for entrepreneurship. A 2018 outage left his Uber team coordinating remotely through fragmented calls and messages, an experience that helped inspire Teamflow, the virtual-office company he founded in 2020. He later described how those coordination problems shaped Teamflow.

Running a startup exposed another problem: incorporation, recruiting, fundraising, scheduling, and communications consume attention before founders can concentrate on their products. Crivello founded Lindy in 2023 to automate this surrounding work. Its initial AI assistant scheduled meetings, drafted personalized emails, prepared users for conversations, remembered preferences, and requested confirmation when decisions were uncertain or consequential.

His thinking about collaboration also shifted. After founding a virtual-office company, he reconsidered remote work and brought Lindy’s team together in San Francisco, emphasizing faster feedback, shared context, and proximity to the city’s AI ecosystem.

The ideas behind Lindy

• AI employees that complete work: Crivello distinguishes generating content from carrying out operational tasks: coordinating schedules, gathering meeting materials, updating records, and following through across multiple applications.

• General-purpose, tool-using agents: Lindy pursues goals through a shared framework that interprets instructions, applies user preferences, and selects appropriate tools. Crivello has also acknowledged the reliability risks of convincing outputs, including an experimental assistant that repeatedly invented nonexistent vocabulary words.

• Multi-agent collaboration: His AI Engineer Summit talk outlined specialized agents that find documentation, divide engineering tasks, handle authentication, test integrations, and prepare pull requests. He presented the workflow as unfinished, not as proof that autonomous software engineering was solved.

• AI employees building other agents: Lindy 3.0 expanded the product into an organizational platform. Agent Builder creates agents from natural-language instructions, while Autopilot gives them cloud-based computers for interacting with software beyond dedicated API integrations.

Crivello ultimately sees agent coordination as a way to lower the barriers to entrepreneurship, allowing smaller teams to pursue businesses that previously demanded greater capital, staffing, and institutional support.