Legendary IBM i is well-known for its stability, reliability, and security.
We know the system is very robust and capable, but it can sometimes feel limiting. This is where IBM i modernization services play a key role. We’ll help you get full value and the startup feel out of this powerful IBM i system.
It’s true that IBM has been investing in the IBM i system, and it has infrastructure advantages that other systems do not have. Also, as your business logic is tried and tested, you only need to upgrade or improve the legacy processes with a few simple yet modern tools and tricks.
That’s where the practices startups follow become worth considering.
What Are the Four Startup Habits for IBM i?
The habits worth borrowing from startups are continuous delivery, faster feedback, documentation, and automation.
- Continuous delivery means changes can be deployed quickly and tested without huge change orders and a slower release cycle
- Feedback means monitoring and alerting, and knowing something has gone wrong before users do
- Documentation means having good documents and procedures for the system and its applications
- IBM i automation means streamlining processes. The fewer manual workflows, the better
IBM i has an integrated system – the database, journaling, work management, object-level authority, IBM i services, test and development environments, and backup software. The box speaks HTTP, which means it connects to other systems.
Habit 1 – Continuous Delivery
Our Senior Engagement Manager for IBM i, Steph Rabbani, says continuous delivery should look like this –
When a developer commits a change > the pipeline builds the change > tests run to check backwards compatibility > the change becomes releasable
So, when a person clicks a button to deploy into a specific library or partition on IBM i, it provides the opportunity to make small changes often instead of bigger ones.
With code in Git and using GitHub, GitLab, or Bitbucket, you can implement continuous delivery without purchasing a package.
A workflow file tells the repository how to build the code. A runner looks for changes, pulls them, and runs the build on IBM i. The pipeline builds from GitHub to IBM i with the click of a button.
Habit 2 – Faster Feedback – Monitoring and Alerting
The best startup rule is to learn what is broken from an alert before the issue comes from a user or a customer.
You can start with a small effort that includes scheduling a job that uses an SQL service like message queue info or system status.
If the SQL query indicates a problem, the system should trigger an alert. The HTTP POST SQL service can send the result to Teams or Slack or send an email.
Another option is the Manzan event gateway or a more robust system like Prometheus and Grafana.
Unit tests are the fastest feedback of all. If code is tested and tests run, they can let the team know that something has broken.
Habit 3 and 4 – Document Everything and Automate Processes
Changes are tracked, and messages usually tie to an artifact, helping make change management part of documentation.
This is where AI can generate documentation by analyzing code. Runbooks and dependencies help developers understand what each program does, what it calls, and which other programs depend on it.
As many senior IBM i programmers are retiring, it’s necessary to get that knowledge documented.
Every manual process is a liability. Hence, we consider automation an important area of our IBM i modernization services.
SQL services can help automate manual runbooks, including profile maintenance, profile checking, backup verification, storage thresholds, and new user provisioning. SQL scripts can also set up and tear down test environments and data.
An MCP server allows an AI agent to connect to the IBM i system. You can start with read-only access and progress to agents doing certain operations.
For example – an agent can investigate a message and propose a solution for a person to read. We can start one thing at a time.
Now, let’s move to building a modern development environment, as part of our IBM i modernization services.
Building a Modern Development Environment
VS Code has the IBM Developer Pack language extension and the DB2 for i extension. See more facts below –
- Developers can use Git and AI extensions
- Git runs natively and for free
- Git can be installed into PASE or locally on a computer
If Git runs natively on IBM i, source can live in the IFS and be compiled straight from the IFS.
Interestingly, once source is in Git, coding standards can be applied automatically. Commercial change-management tools are useful when an audit trail is needed for who promoted what and when.
RPG unit tests can run in the pipeline after a change, and the build can fail if tests fail. SQL scripts can create test environments and repeatable test data.
To get started, focus on the programs that break most often. This provides a practical starting point for improving code quality and introducing automated testing.
AI for IBM i development can further enhance this modern development environment by helping developers understand existing programs, make changes, and support development tasks.
Practical Ways to Use AI for IBM i Development
Effective IBM i modernization services cover one essential component – AI.
AI for IBM i development tools help you make changes or understand programs, including –
- Explaining a program
- Generating or previewing display files
- Writing RPG unit tests
- Converting old fixed-format or System/36 code to newer code
Additionally, the MCP server lets an AI client ask the box questions – what is running, what has changed, and what a program has touched.
No doubt, AI for IBM i development moves the work, but final human verification is needed to ensure changes are making sense. AI-generated tests may pass without covering every scenario, so tests should also account for edge cases and user actions beyond common workflows.
The most important thing to consider here is that AI is not the most expert at RPG, and its implementation of indicators and built-in functions may not always be correct.
Also, AI is a real help when it comes to documentation. You can ask AI
- What a program does
- What it calls
- Which tables are used where
Now, let’s move to IBM i green screen modernization without rewriting the business logic.
IBM i Green Screen Modernization Without a Full Rewrite
Users know the reliability and stability of the IBM i system, but at the same time, they look for a better user experience.
Interestingly, the business logic is already there. What you need to do is make it callable, so new user interfaces can be created over the current business logic.
See three ways to achieve this –
- A quick and easy start – screen scraper
- Creating a new user interface through APIs, with a web, React, or HTML front end on an RPG backend
- Converting an RPG program to a Java or Python app could be a greater effort
AI can analyze the current project, its screens, business logic, and validations, and write a specification.
Once the API is created and started working and delivering JSON, a modern React screen can be created on top of the new API, replacing the green screen.
In this way, the modern screen and existing green screen work together without a big cutover.
It’s important to note that the IBM i modernization strategy becomes successful only if it contains ways to identify and reduce technical debt.
Identify and Reduce Technical Debt
Every organization has technical debt, including a startup. It needs to be measured, tracked, and acted on. Technical debt can include –
- Undocumented business logic
- Dead objects
- Old RPG code
- Duplicated code
- Automated tests
- Manual processes
There may be business logic that nobody knows, and nobody dares to touch.
That’s where AI can analyze code, map it, and create source trees, call graphs, and data dictionaries. Object statistics can show objects that have not been called. A program can be archived, and objects can be deleted.
The Boy Scout rule is to leave each program a little better than you found it. Take out duplicated procedures or make code free-form. Do not pay down technical debt all at once.
Hence, the best strategy is to prioritize tasks > make a plan > work it into sprints over time.
Now, it’s time to consider what we should not copy from startups.
What Not to Copy from Startups?
If you wish to modernize your IBM i without replacing what works, then you should be very particular about the processes that have been running your business for more than 20 years.
- Do not rewrite the stack every three years, skip tests, or fix in production. There is no need to adopt every AI capability immediately. Be deliberate, start read-only, and go from there.
- Start where the pain is within the company. If changes are difficult, make sure developers are using the latest tools. Take one application into Git and create a pipeline that builds and compiles the program on the box. Add unit tests for more confidence in deployment.
- If nobody knows how the box works, use SQL services, active job information, statistics, and AI documentation. If users do not like the interface and think it is legacy, create a REST endpoint over one of the most-used programs, build a modern screen, put it in front of users, and measure the reaction.
You need to measure processes, see how it is working, and adjust accordingly. Take startup operating models, update processes, and use the tools to bring the box out of the legacy perception into the modern startup perception.
An experienced service provider like Integrative Systems helps in the IBM i modernization journey.
Why Choose Integrative Systems for IBM i Modernization Services?
At Integrative Systems, we start by helping you succeed first!
By working as an extension of your team, our IBM i modernization strategy experts help you modernize your system while keeping intact what’s already working.
Our senior engagement manager – IBM i, Stephanie Rabbani, and COO Mike Stein have shared the unique yet powerful tips and tricks to bring modern agility to your IBM i architecture.
Read the page to learn more – https://www.integrativesystems.com/capabilities/ibmi-webinar-2026/
Frequently Asked Questions on IBM i Modernization Services
1. How does legacy Application Modernization AI Reduce Risk?
Ans – AI can help you review IBM i applications, understand dependencies, document set business rules, and plan data-led modernization strategies. Your team knows what needs to be preserved, what needs to be modernized, and what needs to be replaced before the modernization initiative begins.
2. What is IBM AI Project Bob?
Ans – IBM AI Project Bob is a modern-day development environment and intelligent pair programmer. Bob acts as an AI assistant for IBM i developers to help devise complex tasks and coordinate across code, documentation, tests, and the deployment pipeline. It supports diligently in refactoring & modernizing legacy applications.
3. Can an Existing IBM i Application be Enhanced without Replacing it?
Ans – Existing applications can be maintained and advanced with new functionality, integrations, reports, and business changes while retaining established application logic.
4. How Do You Reduce Risk when Changing a Legacy IBM i Application?
Ans – We first understand the requirements and existing application, assess the potential impact, develop the change, test and validate it, and support the release.
5. What are IBM i Modernization Services?
Ans – Through IBM i modernization services, our IBM i experts help modernize established IBM i applications, code, user interfaces, data access, reporting, integrations, and development practices.
With over 20 years of experience in IBM i (AS/400) development, modernization, and integration, Stephanie brings a unique combination of technical expertise, leadership, and customer partnership to the team.
She has led multiple modernization projects across global customers – helping them transform legacy systems into agile, cloud-ready, and future-focused environments.
