How AGITO Global’s Engineers Use AI to Keep Technical Documentation Up to Date

This blog article explores how AGITO Global engineers use AI to reduce documentation work while keeping human judgement, quality and customer value at the centre.

Technical documentation has a timing problem. It is often completed late in a project and can begin to fall behind as soon as the software changes.

For engineering teams, that creates two pressures. Time is spent repeating routine writing tasks, while customers risk receiving guidance that no longer matches the system in front of them. In an automated operation, where people rely on clear instructions during training, testing, go-live and ongoing support, that gap matters.

At AGITO, we are changing how this work is managed. Our engineers are using AI to support the preparation and maintenance of technical documentation, while keeping engineering knowledge, customer requirements and human review at the centre of the process.

Engineering Knowledge Comes First

AGITO engineers work across architecture, interface design, software development, testing, emulation, release management and long-term support. AI is one of many technologies used within that wider engineering process.

Its role is practical. It helps teams organise information that already exists and turn it into a clear first draft for review. The people who understand the system still define how each feature should work, what the user needs to know and whether the final guidance is accurate.

This distinction is important. Technical documentation is useful because it is correct, clear and relevant to the person using the system. Those decisions still depend on engineering judgement.

Using the Software as the Starting Point

Momentum, AGITO’s Warehouse Control System, coordinates equipment, operational workflows and data across automated environments. Its screens, workflows, validation rules and error messages already contain much of the information required for a user guide.

Instead of asking an engineer to recreate all of that information manually, AI can help prepare a structured first draft from approved project sources. These can include application code, user stories, acceptance criteria, interface text and test scenarios.

The work is completed one feature at a time. This keeps each section focused and easier to review. A typical feature guide may explain:

• Where to find the function
• What the function is used for
• The steps required to complete a task
• Common warnings or errors
• Practical troubleshooting guidance
• The screenshots needed to support the instructions

The language is written for the user, not the developer. An operations supervisor working a night shift should be able to follow the steps without needing to understand the software behind them.

From Technical Detail to Usable Guidance

The Notification Handling function provides a practical example. The first draft can explain the difference between individual contacts and email groups, show how to create each one and clarify what happens when a contact is deleted.

The source material provides the technical facts. AGITO’s engineers and analysts decide how those facts should be explained, which warnings are important and what an operator needs to see on screen.

The same approach can identify where screenshots are required and specify what each image should show. This gives the reviewer a clearer starting point and helps keep written instructions and visuals aligned.

Human Review Remains Essential

AI does not decide whether documentation is ready to publish. Every section is checked by people who understand the feature, the operating context and the customer requirements.

The review asks practical questions. Does the guide match the current software? Are the steps in the correct order? Could any instruction be misunderstood? Is a warning needed? Does the wording make sense to someone who has not been involved in development?

This review is particularly important in automation environments, where documentation may support training, system testing, commissioning, user acceptance testing, go-live and day-to-day operation.< AI supports the preparation of the content. AGITO’s engineers and analysts remain accountable for the result.

Keeping Documentation Current

The main benefit is not simply faster writing. It is the ability to keep documentation aligned with the software over time.When a feature changes, the team can review the affected code and project records, prepare an updated draft for that section and complete another engineering check. There is no need to rewrite the full manual or rely on someone remembering every place where the change may appear.

This approach can provide several practical benefits:

• Less manual documentation work for engineers and analysts
• More consistent structure and terminology across features
• A clearer link between the current software and the guidance provided to users
• Faster updates when workflows, screens or validation rules change
• Better support for customer training, handover and ongoing use

It also reduces the risk of different authors describing similar functions in different ways. Shared writing standards help each section follow the same format, level of detail and approach to warnings, notes and troubleshooting.

Supporting Customers Throughout Delivery

Documentation should not be treated as a final task added at the end of a project. It supports delivery from design and testing through to deployment, training and maintenance.

During testing, clear documentation helps teams confirm that the user journey matches the agreed requirement. During commissioning and user acceptance testing, it gives customer teams a practical reference for system functions and expected behaviour. At go-live, it supports training and day-to-day operation. Later, it provides a stronger base for service, software updates and knowledge transfer.

This reflects AGITO’s wider approach to project delivery. Our teams work across software, controls, automation, testing and operational support. Keeping documentation current is part of delivering a system that customers can use, understand and maintain with confidence.

AI as Part of Good Engineering Practice

The most useful applications of AI in engineering are often practical rather than dramatic. They reduce repeated tasks, organise technical information and help experienced people complete routine work more efficiently.

That does not reduce the need for expertise. The quality of the result depends on the standards, context and review applied by the team.

At AGITO, AI sits alongside established engineering practices such as user stories, design reviews, version control, automated testing, emulation and release management. It supports the process rather than replacing it.

The aim is not to produce more documentation for the sake of it. The aim is to give customers clearer information, reduce gaps between the software and the manual, and maintain that quality as the system develops.

What Comes Next

The same method can support other areas of technical content, including interface references, icon lists, validation rules, exception handling and troubleshooting guidance.

Each use case requires the same discipline, supported by approved sources, defined standards, clear instructions and human review.

For AGITO Global, the value lies in applying AI where it can reduce routine work while protecting the accuracy and quality customers expect. The technology may help prepare the first draft, but engineering knowledge remains responsible for the final word.

This article was written with contributions from Dr Rupesh Shet, Head of Software Engineering and Solution Architecture, and Yamuna Kingam, Business Analyst.