An organizational knowledge platform for incubators, built from first principles.
Incubators accumulate hard-won institutional knowledge, then watch it scatter across people, documents, and tools. I designed and built Cortex: a multi-tenant platform that keeps that knowledge inside the organization and lets teams put AI to work on top of it, without betting the company on a single vendor.
Institutional knowledge lives everywhere, which means it lives nowhere.
Incubators and accelerators run on institutional knowledge, and it sits everywhere: in founders’ and advisors’ heads, in documents, chat, email, and a dozen external tools. As organizations grow, that knowledge gets harder to find, harder to reuse, and easy to lose when people move on.
At the same time, teams are wiring workflows around whichever AI vendor they started with, usually without a plan for what happens when costs, capabilities, or requirements change.
Knowledge fragmentation
Critical information is spread across email, documents, chat, CRMs, internal tools, and individual people. When someone leaves or a project ends, it walks out the door.
AI vendor lock-in
Workflows built around a single AI vendor get expensive to leave. As costs and capabilities shift, migrating away becomes its own project.
Repeated work
People keep answering the same questions and rerunning the same analyses, because prior work is too hard to locate and reuse.
The design brief.
A shared agent ecosystem, with each tenant fully isolated.
Cortex is a multi-tenant organizational intelligence platform. Every organization gets a shared agent ecosystem on shared infrastructure, while its data and memory stay completely separated from everyone else’s. Organizations onboard independently, but they all run on the same underlying system.
- Multi-tenant SaaS platform
- Shared compute infrastructure
- Role-based access controls
- Long-running agent sessions
- Tenant-isolated memory
- LLM-agnostic by design
- Shared compute pools
- Persistent agent sessions
- Cloud-native deployment
- Horizontal scalability
- Vendor-independent AI integration
- Tenant isolation throughout
On top of that runs a library of reusable agents that can be deployed across organizations, with administrators deciding which agents each user can reach. Early examples span very different jobs:
Durable memory that outlives any user or model.
The hardest design goal was durable organizational memory that stays secure and tenant-isolated. Cortex uses a hybrid system, so knowledge persists beyond any individual person or AI model.
Relational storage
Structured operational data, permissions, and transactional records.
Document storage
Long-form content, files, reports, and unstructured information.
Knowledge graph
The relationships between people, organizations, projects, conversations, and concepts.
Four ideas the whole design answers to.
Knowledge belongs to the organization
It should stay available even when employees leave or tools change.
AI models are replaceable
You should be able to switch AI vendors without rebuilding your workflows.
Agents are reusable
Build a capability once, then share it across organizations where it fits.
Infrastructure should be invisible
People focus on outcomes, not system administration.
Not modernizing a platform: designing one from scratch.
Cortex is a different kind of engagement from a modernization. It is a new platform designed and built from first principles, bringing together multi-tenant SaaS architecture, knowledge management, agentic AI, memory design, cloud infrastructure, security and access control, and organizational workflow design in one system.
The result is a foundation for preserving institutional knowledge, cutting duplicated effort, and letting organizations adopt AI without becoming dependent on any single vendor.
Carrying production without a platform team?
Start with a 15-minute intro call, or send an email. We’ll talk through where your systems are today and what it would take to make them reliable.
Pick a time that works and it’s booked: a quick call to see whether I’m the right fit for what you’re building.
Pick a timePrefer email? Send a note about your team and systems and I’ll get back to you, usually the same day.
brent@brentmillsllc.com Send an email