About Lyrian
The operating system for the AI-era PMO.
We're building Lyrian because the gap between how AI can help portfolio leaders and how today's tools actually use it is enormous — and closing it changes how programmes get delivered.
Our story
Lyrian was born inside the work. After years of building PMOs from scratch — for teachers' unions, school districts, public-sector transformations, and growth-stage tech — we kept hitting the same wall.
The tools we paid for (Smartsheet, Asana, Microsoft Project, Jira, Monday) were built for the project manager. They were never built for the person three layers above, who needs to know which of forty-two programmes deserves five minutes of their attention this week.
So we did what every PMO does — we built spreadsheets. Lots of them. Reconciliation spreadsheets. Status-roll-up spreadsheets. RAID-tracking spreadsheets. We rebuilt them every quarter. Eventually, we realised that the work we were doing — synthesising, summarising, prioritising — was exactly what large language models had become good at.
Lyrian is what happens when you start from that insight and rebuild a PMO platform from the ground up.
What we believe
AI augments judgement; it doesn't replace it.
Every AI output in Lyrian is editable, attributable, and shows its reasoning. We're not interested in opaque scoring or hidden ranking. The PMO leader is the decision-maker — we just make the path to the decision shorter.
Reports are a side-effect, not a product.
If your team is spending Friday assembling slides, the data model is wrong. Status reports, executive briefings, and forecasts should be generated from the same project graph people work in all week — and updated continuously.
One graph, not eleven dashboards.
Portfolio health, plan progress, resource capacity, financial forecast, governance risk — all of it derives from one connected data model. That's what lets the AI reason across modules instead of working a single slice.
Built for the enterprise, not for the demo.
Roles, permissions, audit logs, RLS-enforced data isolation, exportable everything. Lyrian is designed to survive contact with security review, not just look good in a sales call.
Speed of trust matters more than speed of features.
We ship narrow before we ship broad. Every AI feature has a fallback if Anthropic is down. Every write action is reversible or logged. We'd rather earn a year of confidence than a week of buzz.
Your data stays yours.
Lyrian is single-tenant by workspace and your project data isn't used to train any model. We're an Anthropic API customer — your data isn't visible to them either.
What we're building toward
A world where a portfolio director walks into Monday with their week already triaged — three projects flagged, two recommendations queued, one decision needed. Where the status report writes itself overnight and the human spends the saved hour on the one conversation that changes a project's trajectory.
We're a long way from done. The product you see today is the foundation: a complete PMO platform with AI woven through every module. What comes next is depth — deeper integrations, smarter forecasting, mobile and desktop apps, and an even tighter feedback loop between the AI and the leader using it.
If that's the future you want to help shape, we'd love to talk.
Get on the waitlist
We're onboarding design partners through the private beta. Leave your email and we'll reach out.
Prefer to write a proper note? Use the contact form.