Lyrion has joined the Open Secure AI Alliance, alongside

NVIDIAMicrosoftPalantirIBMAdobe
Read the announcementRead more
Blog

Notes from inside the work.

Field notes on running a company where agents hold the context, take the actions in your tools, and leave a record a person can check.

Article

We raised $10.3M to take the busywork off the company

Article

What an AI operating system actually is

Article

Start with one queue, not the whole company

Article

Approvals are the product, not the friction

Article

Company memory, and why context beats capability

Article

The morning brief: what Muse does before you arrive

Article

Agents and the inbox: what to automate and what to never automate

Article

Audit trails: what your security team will ask for

Article

Escalation done properly

Article

Voice agents in practice

Article

CRM hygiene without the six-month project

Article

What we deliberately do not automate

Article

Measuring agent work honestly

Article

The real cost of a queue nobody owns

Article

How Lyrion works alongside Intercom

Article

The permissions model, explained plainly

Article

Your first week with Lyrion

Article

When an agent is wrong

Article

Documents your agents should be able to read

Article

Does support quality actually go up?

Article

Running a department with agents in it

Article

Data residency and where your work runs

Article

Four integration patterns that keep showing up

Article

Handoffs between agents

Article

Why we show the work

“Muse reads the whole queue before we get in, drafts what it can, and shows its work so we can check it in seconds. Our mornings stopped starting with a backlog.”
Rebecca S.Head of Support