IBM watsonx Code Assistant Portfolio

It allows us to improve automation capabilities within our organization. It address our multiple business automation related concerns and smoothens the workflow. Creating ansible playbooks requires very deep knowledge of autmoation but IBM watsonx Code Assistant for Red Hat Ansible Lightspeed has simplified it to a great extent

We automate multiple tasks at a single go, using their chat style interface, we give large promts and it executes easily

We need to update our applications regularly to keep up with how business is done nowadays and also want to move some of our applications to Java over time because we have a lot of old COBOL code and it's hard to make our applications work well with today's system but it uses AI to change COBOL code into Java, much easier and faster. For specific, well-defined COBOL modules, I can utilize the AI-powered translation to generate a Java equivalent and for critical sections within the chosen programs, watsonx Code Assistant suggests refactoring options. I evaluate these suggestions and apply the ones that enhance code structure and clarity while preserving functionality.

Legacy applications can sprawl over years of development. I've witnessed the struggle to maintain ageing COBOL applications. Watsonx can automate repetitive tasks like generating boilerplate code or basic functionalities and it understands the code and then refactors it and even translates it to Java to migrate applications to a more modern language. It also can automatically generate basic test cases, accelerating the development cycle and as a force multiplier.

It was used in upgrading the version of Java from 11 to 21 in few of the Spring boot applications and custom libraries that were written for internal applications. It doesn't solve business problems exactly as such, it was mostly used for resolving tech debts and improvising existing applications. Since majority of the codebase is having similar tech stack, it has consistent results

It can be majorly used wherever we have standard java (maybe more) applications written in well known frameworks and not too outdated inversion. It doesn't support kotlin that well and is not that versatile. It is not that suitable for custom frameworks built for internal use, general developer productivity like a normal coding agent either

We are using IBM watsonx Code Assistant Portfolio as a coding assistant in software development projects.
We use the basic features such as automatic code generation, code documentation, translation etc.
We use it in conjunction with other AI coding assistants such as Claude.

I think the main features work as advertised, so it is a recommendable product overall. However, just like other AI coding assistants, you need an experienced software engineer in your team to verify, edit and integrate the code generated.
This is not something that will replace software engineers, at least in its current form.

Tool to write complex code of high quality using prompts. So Coder experience level is not that significant but tool enables one to write the code as desired.
Its an AI tool which has modernized the art of writing code and creating platforms for business. The automation has reached another high showcasing technolgical high.

Enterprise modernization, Governance, risks and complaince can be handled in a standard manner.

To modernize the application quickly and efficiently without compromising performance

Converting batch COBOL

Good news customization for Java and intelleju and eclipse
also so it should be useful for developer who write microservices for banking mobile apps and Internet banking

Eclipse and IntelliJ plugin

We POC'd the product for IBM watsonx Code Assistant Portfolio for Z. The use is for Mainframe understanding and modernization. This was to understand how our mainframe components interact and how we can understand what the dependencies are within our different applications. It also provided visual representations that can document what is happening across our pieces. The last IBM watsonx Code Assistant Portfolio piece was done in VSCode to Transform code.

I would recommend for understanding your Mainframe components not for the GenAI piece involved from just my experience. The explanations were not up to the quality we wanted but its deterministic side provided a lot of value for different members of my team. The visuals would be great. I am not sure where it currently stands.

I ran a proof of concept for my Dev team

Discovery of the mainframe landscape of our mainframe and being able to discover all the dependencies

We have a few clients that are running cobol 4.2 and needs to migrate to cobol 6 on many complex codes that they don't have documentation we use IBM watsonx Code Assistant Portfolio to helps them to get what they code makes

I think that it can helps on code debugging an also to improve the actuality of the code