Current direction

We are building the workspace for founder decisions.

Fundador is moving toward a founder decision layer that combines live systems, operational memory, scoped agents, workflows, and simulations so answers can become more grounded, traceable, and useful.

Truthful answers
The product direction is to ground outputs in live data, modeled output, or stated assumptions.
Connected operators
Tool access is being centralized through Composio and MCP-backed capabilities.
Decision workflows
Chat is becoming the front door, with canvas kept for deeper orchestration when the job expands.
Decision workspace
Lead Agent
Intended interaction
Founder question
What breaks if collection timing slips by 20 days while we add two new operators next quarter?
Answer shape

Fundador should compare current capacity, operating assumptions, and workflow load before recommending a hiring or cash-preservation move.

Live data
Modeled output
Assumptions
Inputs
  • Connected apps and operational tools
  • Knowledge and company memory
  • Workflow state and active specialists
Outcomes
  • Scenario comparison and explanation
  • Traceable recommendation
  • Follow-up execution in tools or workflows
What the platform already supports

A founder front door backed by real product layers.

Fundador already has the beginnings of the right stack. The product work now is about making those layers legible, simpler, and more trustworthy at the front door.

Ground on actuals

Pull from live systems, workspaces, and operating records before making a recommendation.

Keep memory durable

Store definitions, assumptions, documents, and company context so the system reasons with continuity.

Model what changes

Use deterministic logic and scenario execution when the question needs more than chat intuition.

Turn answers into work

Route the next action to specialists, tools, workflows, and longer-running jobs without losing the thread.

Truth contract

Every meaningful answer should show its footing.

Fundador should not behave like a generic dashboard or a thin AI wrapper. If it answers a business question, the answer should be grounded in one or more explicit sources.

Live data

Answers should say when they used connected systems or current platform state.

Modeled output

If Fundador simulates or computes a result, that modeled layer should be visible.

Explicit assumptions

If something is missing, the platform should state the assumption instead of hiding it.

Live platform counts

What is already in the system.

These public counts are pulled from the current platform state. They are intentionally operational, not vanity metrics or growth claims.

2
workspaces
shared operating contexts
142
projects
active artifacts across the platform
25
agents
scoped specialists and operators
4
knowledge docs
retrievable company context
0
workflows
automation and orchestration paths
0
active models
available reasoning backends

Plain-language front door

Start with one question, not five dashboards. Chat is the intake surface for founder work.

Canvas as control plane

Keep the existing canvas for orchestration, workflows, and deeper agent configuration when the job expands.

Connected systems

Use Composio for customer app connections and MCPs for scoped tool access, so integrations stay centralized.

Scoped operators

Lead agents can coordinate specialists, but each operator should remain grounded in the tools and knowledge it owns.

Keep the canvas

Simpler front door, deeper execution surface.

Fundador does not need a full design-system reset to feel sharper. The better move is a clearer front door for decisions, with the existing canvas preserved as the place where founders and operators expand the work into orchestrated systems.