Agents should earn trust the way people do: by showing their work
Shoirly exists so that banks can let AI agents do more, because they can see exactly what those agents did and who allowed it.
Why we're building this
AI agents are starting to act inside banks: refunding, reviewing, updating, flagging. Banks are willing to give them more autonomy, but only inside defined limits, and only with proof the limits held.
Today that proof is thin. Certificates show a vendor has controls. Questionnaires record what a vendor says. Neither shows what the agent did on a given day, for a given customer. So security reviews drag on, and good products wait.
We think the fix is simple to describe: sign every action, link it to who authorised it, prove nothing is missing, and hand the bank evidence it can check for itself.
The team
Four founders, based in Dublin.
- Photo to come
Founder name
Role
One line on background and what they own at Shoirly.
- Photo to come
Founder name
Role
One line on background and what they own at Shoirly.
- Photo to come
Founder name
Role
One line on background and what they own at Shoirly.
- Photo to come
Founder name
Role
One line on background and what they own at Shoirly.
Made in Dublin
Dublin sits close to the banks, insurers and payment firms we build for, and to the EU supervisors whose rules shape what they need.
In October 2026 we're taking part in Baseline's Pressure Cooker during Dublin AI Week. If you're there, come and find us.
Talk to the people building it.
Thirty minutes. We'll walk through a sample evidence pack and how it fits your next security review.