Joinable Labs Launches Trusted Knowledge Platform for Agentic AI

Joinable Labs announced the launch of Propagator, the Trusted Knowledge Foundry that turns the raw, unstructured information trapped inside a company’s systems into structured, trustworthy, permission-governed knowledge: the reliable foundation that enterprise AI agents need to do real work. Already powering more than 140,000 AI projects, Joinable gives organizations a way to make their own institutional knowledge usable, and safe, for AI.

Also Read: Infrastructure as Code Governance: Managing Risk in Enterprise Cloud Environments

One of the hardest barriers to useful enterprise AI is doing two things at once: putting decades of institutional knowledge to work for AI agents without giving up control of who’s allowed to see it. From standard operating procedures and technical documentation to security playbooks, policies, and best practices, Propagator turns all of it into AI-ready knowledge that powers agents to reason, decide, and act under the same permissions that govern it at the source.

An enterprise agent is only as capable as the information it can leverage, and most of what a company knows is locked inside PDFs, Word documents, emails, manuals, support tickets, spreadsheets, and chat threads, all created for people, not for machines.

The industry’s default answer has been to copy all of it into a vector database, chopped into disconnected fragments, and leave the understanding for later. That approach centralizes a company’s most sensitive information and strips away the permissions that govern it at the source, pushing the real work to query time. Every time an agent asks a question, it has to search that pile of fragments and piece together the meaning itself, reasoning over text that’s stale, unverified, and stripped of the context that once gave it sense.

Propagator takes a different approach. Rather than dumping documents into an index, it refines them. It extracts the operational knowledge, business logic, and relationships buried in unstructured content and turns them into clean, validated, reusable units of knowledge the company calls Data Cards, each governed by the same permissions as its source. The result isn’t a better chatbot or another agent framework. It’s a trustworthy knowledge layer that any agent can draw on to understand how an organization actually operates.

The distinction matters. Just as a semiconductor foundry doesn’t make the finished phone, but rather the refined chips the phone runs on, Propagator doesn’t build the agents that run your business. It makes the refined, reliable knowledge those agents depend on, so the enterprise can actually trust what its agents do.

“Every enterprise is sitting on an extraordinary competitive advantage: its institutional knowledge,” said Brian Shin, Co-Founder of Joinable Labs. “The problem is that this knowledge was created for people, not AI. Propagator is the layer that turns raw institutional knowledge into structured, trustworthy, permission-governed knowledge an agent can actually act on. We’re not building your agents. We’re building the foundation that lets you trust them.”

Also Read: Infrastructure as Code vs Configuration Management: What’s the Difference?

Refined, not just retrieved

At the core of Propagator is an industrialized refinement pipeline that extracts, structures, and validates enterprise content into Data Cards, each a validated instance of a reusable schema the platform recognizes wherever it appears across a company’s systems. The pipeline is built to know what it doesn’t know: when it isn’t confident how to classify or model something, it routes the item to a human for review rather than guessing, keeping the refined layer honest as it scales. And because the same real-world entity (a vendor, a customer, a project) is resolved across systems that never talked to each other, an agent can reason over a company’s knowledge as a connected whole rather than a pile of disconnected search results.

Governed by design

Because agents act on a company’s most sensitive information, trust is engineered in rather than bolted on. Refined knowledge is held as structured, encrypted Data Cards, kept continuously in sync with the source, and every time an agent retrieves knowledge, the request is authorized against the actual permissions of the person the agent is acting for, evaluated on the exact payload before it’s assembled, and recorded in a tamper-evident audit log. An agent never sees what the human behind it couldn’t. Propagator exposes this knowledge through a native Model Context Protocol (MCP) server and a standard REST API, so any agent framework can connect with identical scoping and auditing.

Proven in enterprise cybersecurity

One of the first enterprise deployments of Propagator was completed in collaboration with one of the world’s largest cybersecurity companies. The project refined thousands of pages of security standard operating procedures, response protocols, and remediation playbooks into structured, permission-governed knowledge, so the organization’s own security remediation agents reason from its approved procedures and best practices rather than generic internet knowledge. Security operations teams use that foundation to analyze incidents, recommend organization-approved remediation actions, and guide analysts through complex investigations, accelerating response while maintaining governance and compliance.

Joinable Runbooks, Joinable’s own security product, is built on Propagator and available for preview: it turns a SOC’s playbooks into AI-ready knowledge that powers the organization’s security remediation agents. Propagator is the foundation for fully customizable agents an enterprise can own.

“The difference is what the agent is reasoning from,” Shin added. “When a security agent draws on the organization’s own approved playbooks instead of the open internet, it’s faster, more consistent, and safe to rely on. Building that trustworthy foundation is exactly what we do.”

A knowledge foundation for every enterprise

Cybersecurity is only the beginning. The same refinement process produces trustworthy, agent-ready knowledge across virtually every enterprise function: IT operations, customer support, human resources, finance, legal, procurement, compliance, engineering, manufacturing, healthcare, and sales. Every department already holds decades of institutional knowledge; Propagator lets organizations continuously turn that knowledge into a foundation their AI agents can act on safely. Joinable Labs believes the next era of enterprise AI won’t be defined by ever-larger foundation models alone, but by how well organizations turn their proprietary knowledge into trusted, agent-ready intelligence.

Write to us [⁠wasim.a@demandmediaagency.com] to learn more about our exclusive editorial packages and programmes.