~/workspace
codefilescontext↻A workspace that stays.
Our workspace design keeps project files and task context together, making continuity a core infrastructure concern.
BUILT FOR AGENTS. DESIGNED FOR POSSIBILITY.
Cloud infrastructure.
Built for AI agents.
Give autonomous workflows a dedicated place to run.
Explore persistent workspaces, isolated execution, and developer-first control.
ISOLATED BY DESIGNPERSISTENT BY NATUREDEVELOPER FIRST
01 / ORCHESTRATEYour agent applicationModels, instructions, and task planning
→02 / EXECUTEDedicated workspaceTools, dependencies, and working files
→03 / CONTINUEPersistent task contextResults, artifacts, and the next step
OUR PRODUCT ROADMAP: PERSISTENCE · ISOLATION · DEVELOPER CONTROL
BRING YOUR OWN STACK
Python
TypeScript
⌘ Bash
◈ Docker
⌁ MCP
01 / THE PLATFORMLESS INFRASTRUCTURE. MORE POSSIBILITY.
Not another session.
A place to keep going.
Agents do more than answer questions. They write code, use tools, and return to unfinished work. Their environment should be built for that.
01.1
agent A
your agent
agent B
Room to experiment.
Dedicated environments are at the heart of the design. Give agents a place to explore without sharing your local machine.
●Ⅱ↗RUN PAUSE RESUME
01.3Work at your own pace.
Pause-and-resume is a core roadmap priority: preserve useful state between periods of execution rather than rebuilding context for every task.
02 / DEVELOPER ARCHITECTURE
Your application.
Your orchestration.
A dedicated execution layer.
Our architecture direction keeps task planning in your application while organizing tools, files, and execution around a persistent workspace.
Explore the developer architecture ↗01
Define the task
Establish inputs, permissions, and expected outputs.
02
Execute within a boundary
Separate agent-generated operations from application infrastructure.
03
Preserve what matters
Keep artifacts and context available for subsequent steps.
03 / WHAT WILL YOU BUILD?SMALL STARTS. OPEN ENDS.
For agents with
things to do.
From the first prototype to the next big idea.
Explore the workflows we’re designing for.
Example workflow: clone a repository → run tests → save changes → resume tomorrow.
THE INFRASTRUCTURE LAYERExplore the architecture ↗
From one agent
to an entire workflow.
A product direction that separates orchestration from execution. Keep your application in control while each agent gets a workspace of its own.
01 / YOUR APPLICATIONAgent orchestratorYour model · your framework · your logic
→02 / CONTROL LAYERHamantouWorkspace lifecycle · policy · context
→03 / EXECUTION LAYERDedicated workspacesCode · files · tools · artifacts
REFERENCE ARCHITECTURE · OUR PLATFORM DIRECTION
04 / A FEW GOOD QUESTIONS
Good to know.
What is Hamantou building?＋
Hamantou’s product roadmap focuses on cloud workspaces for AI agents: persistent task context, isolated execution, and developer-controlled workflows.
Who is the platform designed for?＋
Our target users are teams building coding agents, research assistants, operational automation, and agent-native products.
How does the architecture fit an existing stack?＋
The design separates orchestration from execution. Your application retains task planning and model selection; the workspace forms the boundary for files, tools, and task state.
How can investors and partners connect?＋
Contact hello@hamantou.com to discuss our company, infrastructure thesis, and roadmap, or call +1 (302) 687-4164.
LET’S TALK INFRASTRUCTUREBring us your next
Bring us your next
agent workflow.
Discuss product fit, integration requirements, or an enterprise deployment.