About Floyd
Your agents deserve
a computer that never turns off.
The Problem
Developers build agents with Claude Code, Cursor, and dozens of other tools. The agents work. They write code, search the web, make decisions. But they only run while your laptop is open.
Close the lid and the agent stops. Open it tomorrow and it has forgotten everything. Want it to run overnight? You're stitching together serverless functions, fighting cold starts, losing state between sessions, and building your own scheduling infrastructure.
The tools for building agents are incredible. The tools for running them don't exist. Floyd is the missing piece. Write your instructions and your agent runs in a persistent computer that never shuts down.
What we believe
01
Write instructions, not infrastructure
Tell your agent what to do in plain text. Floyd gives it a persistent computer with bash, file system, and web search. No Docker, no YAML, no config files.
02
A computer that never turns off
Each agent gets its own persistent environment. Install packages once, create files, build databases. Everything survives between runs.
03
Agents that remember
Every run picks up where the last one left off. Files, databases, context, all preserved. Your agent builds knowledge over time instead of starting from scratch.
Get in touch
Have questions?
Whether you're evaluating Floyd, want to discuss your use case, or just want to talk about the agent infrastructure space.