The IBM i you know.
The agents you don't.
One lens for both.
OneLens4i helps you govern what AI agents can do alongside your IBM i. Set their limits, keep your business rules on IBM i, and trace each request from the agent to the IBM i job and its Db2 for i record. Written by a CTO from a decade inside a regulated IBM i company, and free with your pilot.
Reggie Britt · Founder and CTO · Forty years on IBM i
Running today on our own agents and our own IBM i: 15 agents, every record verified, and calls traced to the IBM i job. See it running →
Every IBM i vendor sells a tool.
We've had to govern them.
I spent a decade inside a regulated IBM i company and served as CTO of a billion-dollar consumer finance business. In a company like that, governance isn't a side job. It is the job: every change answers who decided, on whose authority, and whether you can prove it.
IBM i, first.
We know the box and the code, and we watch the agents beside it as one system, so your team isn't learning a second world alone.
A board's view of AI.
Security, governance, return and people, in that order, because that's how the questions come. Forty years on the platform, from System/36 to IBM i.
People and organization, built in.
Agents change how people work. We plan that with the build, not after it: who governs each agent, and how roles change.
How we bring people along →
We run our own company this way, on a system we built ourselves.
Reggie Britt · Founder and CTO · Forty years on IBM i · See it running →You keep everything built for you. Your logic stays on IBM i, your know-how is written down where your people can read and change it, and the record of every decision is yours. Change your tools next year and all of it stays. Ask any vendor, including us: what would it cost to change tools next year? What you keep →
Build on what you own.
Seven moves, one lens.
OneLens4i sets out what must be true before any agent touches your IBM i, and how to watch the agents and IBM i as one system. Your source, your data and your way of doing business are already paid for; OneLens4i builds on them in seven practical moves. The white paper is a free download, and yours to keep. Run a pilot, and OneLens4i comes with it.
- 1 · Build on what you have. Reach your core systems without rewriting them. →
- 2 · Choose tools with care. Never rent what you can't afford to leave. →
- 3 · Put governance in place. Four gates, and a named owner for every decision. →
- 4 · Put security in place. Every agent gets its own identity and limits; the baseline is assessed and fixed before production use. →
- 5 · See what's happening. One view across the agents and IBM i, from the request to the record. →
- 6 · Set one clear objective. One workflow, one business outcome. →
- 7 · Pilot with financial targets. Agree the numbers before you spend. →
OneLens4i shows the shortest safe path: one workflow, one agent, the four gates and the tools you already have. It's one your own people can maintain.
Scenario: Normal day. The animated floor shows on wider screens; the lens below follows each scenario.
An agent went shopping.
IBM i made the call.
Alder & Pine is a home furnishings store. Silver Lake is the lender behind its financing, and its credit decisions run on IBM i. A shopper sends an AI agent to buy for them. Here's what happens.
-
Step 1 · the agent's side
The agent shops
A shopper sends an agent to furnish a room. Before it can do anything, the store checks which agent it is, who sent it and how far it's allowed to go.
SecureEvery agent has its own identity and its own limits. -
Step 2 · IBM i's side
IBM i decides
The shopper picks financing. Silver Lake's credit decision runs where it always has: the lender's own rules, on IBM i. Only an ID and the amount cross, no Social Security number and no date of birth. Nothing on IBM i was rewritten. A door was put in front of it.
SecureData stays inside its boundary.840 msa real decision from IBM i, on test traffic -
Step 3 · the agent's side
A person commits
The agent builds a $4,980 cart. With $1,899 already spent, that would break the shopper's $5,000 limit. The store says no. The shopper raises the limit and makes the call.
GovernedThe agent proposes. In this workflow, a person approves the commitment. -
Step 4 · across both
It's on the record
The decision is written once to the record, with who asked, what was decided and on whose authority.
GovernedEvery decision traces to a person.275 of 275automated checks pass in the demoIncluding the ones that must fail: no key, a forged key, over the limit, expired permission. Measured live at silverlakefinancialgroup.com ↗
Swap in your own decision: credit, pricing, an allocation, an order release. Your rules stay on IBM i, the agents come to the door, and a person makes the call. The pattern is the same.
Alder & Pine Home and Silver Lake Financial Group are fictitious demonstration companies, built and run by Pegasus Source. Nothing is really sold or lent.
Start free. Move one workflow.
Hand it over.
- Free
The white paper and the Read
The OneLens4i white paper is a free download in your workspace, with the technical companion coming soon. The 3-minute Read gives your score and a gap to look at first.
- Paid assessment
Gate check and baseline scan
We check your IBM i against the four gates and scan its security baseline. Scoped with you.
- Pilot · 90 days
One workflow, moved to agents
You sign the measures first, and OneLens4i comes with it at no additional cost. We build, watch both worlds with you, then hand it over. Ongoing oversight and support are optional.
Every agent earns its placeEach one carries a business case, measured on the decision it serves: its cost, its time, and how often a person steps in. You sign the measures before we build.
People and organization move with every step: who governs each agent, and how roles change, is decided as we build. At every step, what we build is yours. We stay because we keep it improving, not because you can't leave.
This is what you open.
Free, with your email.
The OneLens4i Governance view and white paper, plus the tools for finding where agents pay off, where your people stay in charge, and what each agent should return. Pick one to look inside. The screens show Silver Lake, a fictitious lender; yours open with your own workflows.
What moved on the line this week.
Built In.
The best of the daily log, every two weeks. One idea, one proof, one next step, three minutes. AI-native IBM i, and the agents beside it, from Pegasus4i.
