Core transaction (CICS)
Read and change, per the contractSOAP web services with mutual TLS, or JSON through z/OS Connect. The contract you publish fixes which program is called.
- Mutual TLS
- Program fixed by the contract
- Each person's account
Email, tickets, your APIs, databases and the core. The agent only uses what you publish and its rules allow. Nothing to rewrite.
Agents run only what your team publishes and each agent’s rules allow: executable use cases and Client Bridge actions. The rest of the system is understood and documented.
Search for your technology: you'll see what Reelease understands about it and where an agent can act on it.
What an agent can do on a connected system is set by its rules.
Try:
43 systems and tools
Reads.cbl .cob .cpy .copy .pco .bms .csd .jcl .proc .prc .pli .pl1 +3
Reads.pkb .pks .sql .pls .psp .fmb .olb .pll .mmb .rdf .pc .pcc
Reads.rpgle .sqlrpgle .rpg .clle .clp .pf .lf .dspf .prtf
Reads.nsp .nsn .nsm .nsl
Reads.pbl .srw .srd .dfm .pas .dpr .frm .bas .cls .vbp
Reads.aspx .ascx .master .svc .asmx
Reads.java .kt
Reads.yaml .json .ts .py .rb .erb .graphql .gql .tsx .vue .svelte
On every route, each step declares whether it reads or changes data, your rules check it first, and changes wait for whoever you choose. Everything goes into the run record.
SOAP web services with mutual TLS, or JSON through z/OS Connect. The contract you publish fixes which program is called.
The operations you already expose: OpenAPI, Java, NestJS, PL/SQL web, GraphQL, .NET WCF and ASMX services, IBM i programs and Natural subprograms through EntireX. Every call comes from the contract you publish.
Statements and procedures your admin publishes, inside your network. Each one is tested without leaving changes, and the tunnel goes out from your network with no inbound ports.
The Reelease Agent works in the operator's browser and desktop, with their consent. Every action that changes something waits for a person.
Each person connects their work account, and the target system sees that person, not a technical account.
Your admin installs the Reelease app, and each agent uses only the repositories or channels you assign.
Tell us what you run and we'll tell you how we'd work with it: what we can read from its code and where an agent could act.
Depending on the case, we stand up a copy of your system with test data and the same behavior, so you can publish and test executable use cases before touching the real one.
Missing a system? Tell us which one
Trademarks and logos belong to their respective owners. They are shown to identify integrations and do not imply affiliation or endorsement.
You publish executable use cases on your production systems, from the core to your APIs, and the agent runs them with pre-run checks, rules it cannot skip and a run record.
Laura looks for the template she needs: resolve a held transfer.
She adds the transfer ID and the agent prepares the plan before touching anything.
The signed plan reads the transfer and its hold, and leaves the retry for last.
The reads go ahead on their own. The retry changes data and waits for someone else.
Laura can’t approve her own task: she can only reject it.
María sees the change with its data and approves it, with a reason that stays on record.
The reads and the change finish, and Laura gets the result in her chat.
The admin opens the run record: who launched it, who approved it and why.
With the plan, the steps and every decision. And you can download it.
And in technical verification, fingerprints of the approved plan and the change’s exact values.
Every task, done to its plan, with its run record.
With Bridge Admin, your admin publishes statements and stored procedures from SQL Server, Oracle or PostgreSQL as actions, inside your network. Each one is tested with no changes kept, and you can roll each action back to its previous version.
Whether an action changes data is set by its contract, never by the model. When in doubt it counts as a change and, unless you pre-authorize it, waits for the person you choose.
Agents work today on what you already have. When you replace a system, you switch the connection and the work of agents and people stays the same.
Agents run your executable use cases on the core as it is, in COBOL and CICS if needed.
You choose what to replace and when. What Reelease has understood and documented about your system, business rules included, is your starting point.
Your admin points the connection to the new system and tests it first. Agents, executable use cases, approvals and run records stay as they were.
And if you rebuild a piece on a new system, a database or another service, what Reelease already understands about the current piece (screens, business rules, flows) is what you rebuild it from.
| What’s included | What it isn’t |
|---|---|
| Analysis of your systems, from Java and APIs to COBOL and CICS, with their documentation and use cases, to operate and replace them | A full rewrite before you can start operating |
| Publishing executable use cases on the core, your databases and your APIs | An agent with unrestricted access to every batch job or the mainframe |
| Runs with pre-run checks, rules, the approvals you define and a run record | An isolated demo, disconnected from production |
| Replacement one piece at a time, without agents or people changing how they work | Running anything nobody has published |
| An automatic migration to the new system |
No. In Microsoft 365, Gmail, LinkedIn and Atlassian each operator connects their own account, with scoped permissions; in Microsoft 365 and Gmail, only their work account, never a private one. GitHub and Slack are connected by installing the Reelease app in your GitHub organization or Slack workspace, and each agent’s rules decide what it can do with them.
An admin installs the Reelease app in your GitHub organization and picks each agent’s repositories. Access is limited to those repositories and never reaches the operator’s computer. If you uninstall or suspend the app, access stops.
The Reelease app is installed in your workspace and each agent only reads and posts in the channels you assign to it. Its access is stored encrypted and never reaches the operator’s computer. Revoke it and access stops.
The pre-run check catches it: if a required account is missing, the run doesn’t start.
Yes. We analyze COBOL/CICS, PL/I, IBM i, Natural, PL/SQL, Oracle Forms and Reports, .NET Framework, desktop applications, Java, Python, Ruby, Node, OpenAPI and React front ends. The agent runs what you publish through REST APIs, CICS transactions, PL/SQL web or a Client Bridge in your network (SQL Server, Oracle or PostgreSQL).
Yes, through the Client Bridge or the service you publish. The connection to your environment is validated during the pilot.
Yes. Depending on the case, we stand up a replica of your system with test data and the same behavior. You publish and test executable use cases on it and, once you’re satisfied, the connection moves to the real system.
Tell us which one. Each new family of systems is one more connection path: executable use cases, rules and run records don’t change. And if you would rather rebuild it on a database or another system, what Reelease understands about the current one is what you rebuild it from.
They keep working as before. Each executable use case calls a connection, and the connection points to the system. Your admin points it to the new system from the panel and tests it first; if the new system needs different data, you adjust the connection. You replace one piece at a time, at your pace.
No. Only what is published and allowed runs. A batch job without a published executable use case doesn’t become executable on its own.
We’ll show you Reelease on one of your cases: what it understands, the plan an agent would propose and what you’d sign off on.
You choose the cookies
We use technical storage to remember your choice and, only if you accept, analytics cookies (Google Analytics) to understand how the site is used. You can change your choice at any time. Learn more in our cookie policy.