Kyvvu

Your industry

Your sector isn't listed. Your agents are.

Whatever your company makes or sells, its agents do the same kinds of work: paying invoices, answering customers, handling employee data, helping IT. Kyvvu's policies are about what an agent does, so they apply to yours as they are.

hr-agent · task 2f61kyvvu engine · in-process
  1. step.messagePOSTsmtp/externalblock
Blocked: outbound send after personal data was read in this taskTry it: click the highlighted step to take it off the path

The problem

The same risks, in every company.

  1. 01

    Agents touch money

    Invoices, payments, refunds and credits: a payment an agent shouldn't have made is hard to get back.

    payments

  2. 02

    Outside content gets in

    Emails, documents and web pages from outside can carry an instruction the agent wasn't meant to follow.

    prompt injection

  3. 03

    Personal data is everywhere

    Customers, employees and suppliers: the data an agent needs to do its job is the data that must not leave.

    personal data

In practice

Agent jobs you'll recognise.

An invoice is paid after finance approves

A finance agent matches an incoming invoice to an order and goes to pay it straight away.

finance-agent · task 7b30kyvvu engine · in-process

Illustrative scenarios. Each one is decided by a single policy from the open policy library, as the engine evaluates it.

How Kyvvu helps

One engine, whatever your sector.

  1. 01

    Judges the whole path

    Every action is checked against everything the agent did earlier in the task, and blocked before it runs if the sequence is dangerous.

    allow/warn/block

  2. 02

    Standard packs to start from

    OWASP agentic, the EU AI Act, GDPR, DORA and MiFID II, healthcare and data exfiltration packs from the open library, next to your own rules.

    open policy library

  3. 03

    One policy, every agent

    The same rules for plain Python, LangChain, LangGraph, CrewAI and Claude Code agents, in the cloud, on-prem or air-gapped.

    Python/LangChain/CrewAI/Claude Code

  4. 04

    Hosted or yours

    Use platform.kyvvu.com, or run the platform in your own infrastructure, on-prem or air-gapped, on Enterprise & Public.

    hosted/self-hosted/air-gapped

  5. 05

    Rules people can read

    Policies are short YAML files in your own git, reviewed and versioned like code. Every verdict in the trace names the rule that decided.

    policies as code

  6. 06

    Nothing breaks on day one

    Only critical policies block; high, medium and low ones warn. Start a new rule as a warning and make it critical once you trust it.

    shadow/suggest/enforce

Questions

Common questions.

01

Is Kyvvu only for regulated sectors?

No. The policies are about what agents do, not about which regulator you answer to. Banks and insurers use the same engine as any other company.

02

Where do we start?

With one agent. Run it under the standard packs from the open library, see what they would warn about, then add your own rules.

03

Do we have to change how our agents are built?

No. Kyvvu works with Python, LangChain, LangGraph, CrewAI and Claude Code agents, and with agents in any language through the local engine.

Get started

Every agent has to ASK before it acts.

Start with one agent

See it on your own agents

A 30-minute demo with someone who builds Kyvvu: your agent, your rules, and the policy that decides each step.

See it on your agents

Bring one agent to a demo

A 30-minute demo with someone who builds Kyvvu: your agent, your rules, and the policy that decides each step.

Want to see other teams and sectors? All use cases.