Hey everyone! This is Paulina & Matthew from Agentic Fabriq.

TL;DR: AI agents are entering production — but identity systems weren’t built for them. Agentic Fabriq is the identity and governance layer built specifically for AI agents.

Our launch video: https://youtu.be/6_iIijfy0Mg?si=YMEZHYwgIa8obZ8T

The Problem

Agents are moving fast. Security isn’t.

Teams are shipping AI agents that call tools, trigger workflows, access internal data, and act on behalf of users. But the identity layer underneath them was designed for humans logging into apps — not autonomous software acting across systems.

Today, teams hardcode tokens into workflows, drop user identity once an agent runs, and re-implement OAuth for every integration. Security teams lose visibility into what agents are accessing and doing, and no one can confidently answer a simple question: who did this agent act for?

OAuth, IAM, and IdPs were built for humans and apps. They were never designed for agents calling tools, calling other agents, and operating autonomously across environments. Engineers get stuck rebuilding auth, IT struggles to govern permissions, and security teams are left without clear audit trails.

What We Built

Agentic Fabriq is the identity, governance, and visibility layer for AI agents.

We sit in the middle of every agent’s calls and manage both agent identity and user identity. We handle token exchange and authentication, enforce least-privilege access, and enable fine-grained, per-user permissioning for every agent. Every action is logged centrally for audit and compliance.

Instead of wiring OAuth, scopes, and policies for every agent and every workflow, you call us once.

You connect your tools, register your agent, define the permissions — and we handle identity propagation, authorization, and logging.

Here’s a demo of our permissioning feature with OpenWebUI: https://youtu.be/ayiZ_7d7jCY?si=M3UK8d1B0tvt-8wE

What This Enables

• Per-user agent access
• Agent + user pair permissioning
• Full visibility across tools and workflows
• Instant revocation and policy control
• Compliance-ready audit trails

We make it easy for companies to make their agents secure.

Why Now

Every major computing shift forced identity to evolve.

SaaS required SSO.

Cloud required IAM.

APIs required OAuth.

Agents are the next shift – and they need their own security platform.

We’re building it.

Our Ask

If you’re building internal agents, shipping AI into enterprise, or thinking about agent governance and security, we’d love to talk.

CONTACT

• email us: founders@agenticfabriq.com
• book a meeting: https://calendly.com/paulinaxu/agentic-fabriq-meeting
• website: https://www.agenticfabriq.com

TEAM

Paulina and Matthew met @ MIT admit weekend and have been friends since. They dropped out before their second year to work on Agentic Fabriq

Paulina was studying AI + Physics and doing CS/ML research at MIT Kavli Institute, MIT Haystack Observatory, and INAF Padua.

Matthew was studying AI + Math and doing CS/Econ research at MIT Futuretech Lab and Computational Biology research at MIT CSAIL.