The policy-compiled operating system for enterprise AI agents

Define your rules once. pactflow compiles them into every agent, model, prompt, integration, and permission you ship so teams launch AI without bypassing compliance, IT, or risk.

SOC 2 Type II · Deploys in your VPC · No agent ships un-governed

pactflow control plane · acme-corp
312Agents governed
100%Policy coverage
0Un-reviewed changes
governed_deployslast 14d
support-agent · gpt-4o approved
billing-agent · prompt v12 in review
data-agent · new tool blocked
intake-agent · perms approved

One policy. Every surface an agent touches.

Agents

Register, version, and gate every agent in one control plane.

Models

Pin approved models and routes; block the ones risk hasn't cleared.

Prompts

Treat prompts as code diffed, reviewed, and rolled back on demand.

Integrations

Allow-list the tools and data each agent can reach nothing more.

Permissions

Bind scopes to identity and policy, enforced at every call.

Trusted by platform & risk teams at

NorthwindMeridianHalcyonAtlas BankVeratechLumen Health
01

Compile policy once

Write your guardrails as code once. pactflow compiles them into every agent surface, so the rules can't be skipped, forgotten, or copy-pasted wrong.

Try a live example
# policy.pact compiled into every agent
policy "support-agents":
    models:   ["gpt-4o", "claude-opus-4"]
    tools:    allow("zendesk", "kb.read")
    pii:      redact("email", "card")
    change:   require_review("risk")

# ship it guardrails travel with the agent
$ pactflow compile && pactflow deploy
✓ 312 agents recompiled · 0 violations
02

Deploy with guardrails attached

Every agent reaches production with its models, integrations, and permissions already bound to policy. No side doors, no shadow deployments, no agent running outside the control plane.

See the platform
agent-os · deploy
identity & scopes resolved ok
model allow-list enforced ok
tool permissions bound ok
PII redaction active ok
audit stream → SIEM live
03

Govern change in real time

A new model, a tweaked prompt, a fresh integration every change runs through change-control review before it can reach a single user. Approvals, diffs, and rollbacks, logged for audit.

Explore change control
change-control · review #4471
billing-agent · prompt.diff
-  refund up to $500 without approval
+  refund up to $50; escalate above

reviewers: risk ✓   finance ✓   eng pending
status:   awaiting 1 approval not yet live

Designed for everyone who owns the risk

Ship agents without waiting on a review queue

Policy is already compiled in. Build, test, and deploy against the same guardrails production uses locally, from day one.

  • SDKs for Python, TypeScript, and Go
  • Local policy emulator fail fast before you ship
  • Prompts and tools versioned alongside your code
Read the docs
from pactflow import Agent
agent = Agent("support", policy="support-agents")
# guardrails enforced on every call
agent.run("refund order 8821")

One control plane for every agent in the org

See what's deployed, what it can touch, and what changed across teams, clouds, and vendors from a single dashboard.

  • Central registry of every agent and its scopes
  • Org-wide model allow-lists and rate policy
  • Self-host in your VPC; SSO, SCIM, and RBAC built in
Tour Agent OS
312agents
47policies
9teams

Evidence the auditors actually want

Every decision an agent makes and every change a human makes is logged, attributed, and exportable. Prove control without chasing screenshots.

  • Immutable, signed audit trail streamed to your SIEM
  • Change-control approvals mapped to policy and owner
  • SOC 2, ISO 27001, and EU AI Act reporting templates
See change control
2026-06-26 14:02 · model swap approved · risk
2026-06-26 11:40 · tool added blocked · policy
2026-06-25 17:15 · prompt v12 in review

Risk teams stopped saying no.

“We went from a six-week security review per agent to a same-day approval. pactflow made the guardrails the default instead of the afterthought.”

DK
Dana KhouryVP Platform Engineering, Meridian

“Finally, one place where I can see every agent and exactly what it's allowed to touch.”

RM
Raj MehtaCISO, Atlas Bank

“The change-control log is the first AI evidence our auditors didn't push back on.”

SL
Sofia LindqvistHead of GRC, Lumen Health

“Policy as code clicked for our engineers in an afternoon. Now it travels with every deploy.”

TO
Tomas OduyaStaff Engineer, Veratech

“We ship agents faster and sleep better. That combination didn't used to exist.”

AW
Amara WellsDirector of AI, Northwind

About pactflow

We started pactflow after watching great AI projects die in review not because they were unsafe, but because no one could prove they were safe. So we built the layer that compiles your rules into every agent and keeps the receipts. Govern once; ship forever.

Read our story

Featured blueprints

Start from a policy that already passed review

Support

Tier-1 Support Agent

Refund limits, PII redaction, and escalation built in.

Finance

Invoice Reconciler

Read-only ledgers, dual approval on any write.

Healthcare

Intake Assistant

HIPAA-scoped data, no model leaves the VPC.

Sales

RFP Drafter

Brand-safe prompts with a human-in-the-loop gate.

IT

Access Triage Bot

Least-privilege scopes, every grant logged.

Access the control-plane playbook

A field guide to launching AI agents that compliance, IT, and risk sign off on the first time.

The Agent Governance Playbook

38 pages · 12 reference policies

Get the playbook

Free. No agent emails you back.

Stay connected with the governance community

Practitioners shaping how enterprises ship AI responsibly share policies, compare notes, and trade war stories.

Ship the agents your risk team will approve.