Irys Enterprise
Inside your perimeter, in days.
The same infrastructure behind Irys One, deployed air-gapped, on-prem or in your VPC. Any model, swapped when a better one ships. Your workflows, permissions and record, across every team.
What a deployment includes
One system, from the first team to the whole company.
Not a tool bolted onto your stack. The reasoning core runs where your data already lives and takes on your rules.
The Irys reasoning core: persistent state, verification, permissions, model routing and deployment controls
Your choice of model, local or approved external, swapped when a better one ships
Connection to your systems, documents and records, where they already live
Your workflows, encoded once and reused by every team that follows
Permissions and approval gates matched to how your institution signs off
Discovery, integration, testing and training
Hosting, model access, monitoring and support
Where it runs
Your institution, Irys and any model, all inside your perimeter.
Reads, reasons, checks, waits for approval, and remembers. Every job leaves state behind for the next one.
How it goes
Days to scope and deploy, not quarters.
Every deployment follows the same four stages, and each one ends with something your team can use.
Evaluation
We scope one real workflow with your team and agree what success looks like.
Paid pilot
Irys runs that workflow inside your environment, on your data, against the agreed measure.
Private deployment
The pilot becomes production, under your permissions and on your record.
Expansion
The next team starts on the same core, with what the first team built already in place.
What is reused
Deployed like services. Compounds like software.
What we build for one deployment goes into the core, so every deployment starts further ahead than the last. Your data never moves.
Inside your perimeter the same thing happens with your own work. Each job leaves its state behind, so by job 100 the system starts from accumulated context and reviews only what changed.
Same core. Start with One, move to Enterprise when the rest of the company asks.
In production
Built for legal. Taken everywhere by customers.
Legal was the first application and the hardest test. The same core now runs well beyond it.
- LegalLive
- BankingLive
- IndustrialLive
- ConsultingLive
- DefenceIn pilot
- InsuranceIn pilot
Proof
Numbers, with the method attached.
Irys on a low-cost model, against the legal AI leader's own agent, on the leader's own public benchmark. 2,010 tasks, $4.64 per task.
To scope and deploy inside an institution's own environment.
Every job makes the next one better. At a fintech deployment, review effort per matter is down 90%.
Strict all-pass: a task counts only when every rubric criterion passes, with no partial credit. Legal Agent Benchmark v1.0. Irys ran the full public set of 2,010 tasks across 27 practice areas on a low-cost off-the-shelf model, with no fine-tuning and every task starting from empty state. The benchmark owner reports its own agent on a private held-out split of roughly 1,200 tasks, so the comparison is across the same benchmark family rather than identical task instances. Methodology, caveats and every output are public on GitHub. See the repository.
“It's become our system of record.”
SSO, SCIM, audit logs, MSA, DPA and security review support are covered on Enterprise governance and Security.
Talk to us about Enterprise.
Tell us the workflow and the environment. A member of the deployment team replies within one business day, and the first conversation is about scope, not a pitch.