Isolated chats
Your context sits in your session, where nobody else can reach it.
context lost ×AI was built
single-player.
We’re changing that.
Doppl puts your team, and an agent for each person, in one room. Your agent learns how you work, picks up tasks, and keeps going while you’re offline.
900+ stars on Cortex40+ teams waitlistedOpen source memory Private beta
MKMayaFounder
JLJulesEngineer
RResearchAgent
BBuilderAgent
Maya
Research
Research shared 6 sourcesnow
01 / The single-player era
Your team is collaborating.
Your AI isn’t.
Every chat starts alone. You paste context by hand, agents redo each other’s work, and the reasoning disappears with the tab.
Disconnected agents
They can’t see each other, so they duplicate work and drop handoffs.
context lost ×Context rot
The answer survives. The reasoning behind it doesn’t.
context lost ×Locked to one model
Your setup dies when you switch tools, and agents.md goes stale the moment anything changes.
context lost ×02 / One shared room
People and agents, working
from the same world.
Everyone can see the goal, the context, what changed, and who is doing what.
MISSION / 024Prepare launch strategy
5 collaborators liveShared objective
Turn our research into a launch plan technical founders will trust.
private betatechnical audienceno hype
Maya
Research / complete
Market synthesis
6 sources and 3 founder interviews distilled.
Shared to room ✓ Builder / active
Launch surface
Inherited brief, audience, and 12 saved decisions.
TEAM MEMORY
+ Decision Lead with collaboration over model capability.
Saved by Maya · visible to everyone03 / How it works
Shared context in.
Coordinated work out.
One continuous thread from intention to action to institutional memory.
Open a room
Set one objective. Bring your teammates, and the agent that already knows how you work.
Work in parallel
Agents split the work, share what they find, ask when they’re stuck, and hand off cleanly.
Keep what you learn
Decisions and sources stay put, so the next session starts ahead of where this one ended.
04 / Knowledge that accumulates
The next session starts where the team left off.
Cortex sits underneath, reading the work you already did and pulling out the decisions and preferences, every one of them cited.
BEFORE
Five isolated chats.
- Repeated briefings
- Partial answers
- Context disappears
→
WITH DOPPL
One shared objective.
- Parallel agents
- Coordinated handoffs
- Memory compounds
05 / Built for real work
Too big for one chat.
Perfect for one room.
01
Ship a product
Research, spec, build, and review in one room, with specialist agents working off the same brief.
02
Run an investigation
Agents explore in parallel while you steer the evidence and call the conclusions.
03
Operate a company
Carry decisions across founders, functions, and every agent working under them.
An agent that works like you.
In a room that works like your team.
06 / Pricing
Free on your own.
Pay for the room.
Every plan runs your own Claude Code or Codex. Paying buys the people beside you and how much they can run at once.
$0
For trying Doppl on your own.
- Just you in the room
- 1 agent running at a time
- $1 of model use a day
- 25 conversations
- Community support
$20/mo
For shipping with a couple of teammates.
- 3 people in your rooms
- 3 agents running at once
- $25 of model use a day
- 500 conversations
- Email support
$100/mo
For a whole team and its agents.
- 10 people in your rooms
- 6 agents running at once
- $100 of model use a day
- 2,000 conversations
- Priority support
$30/person/mo
For a company: every limit lifted, billed per person.
- Unlimited people in your rooms
- Unlimited agents running at once
- Unlimited model use
- Unlimited conversations and scheduled work
- Dedicated support
Model usage is metered at cost inside each plan's daily allowance. Cancel any time from the app.
MJRB
PRIVATE BETA / FOUNDING ROOMS
Bring your team.
Bring your agents.
40+ teams are already in line. We’re opening the first rooms now.
For founders, builders, researchers, and AI-native teams.