Problem Statement

Customers have extensive needs to modernize their mainframe to cloud because of 1/ High cost of maintaining mainframe 2/ Diminishing knowledge of mainframe language such as COBOL on the market 3/ Significant benefits of agility and economy of cloud service. Since the architecture of mainframe is different from modern cloud structures, mainframe modernization usually takes multiple steps. Depending on the patterns the workflow could be different. I was working primarily on the pattern Refactor which includes the following steps: 1/ Understand code based on functionality 2/ Decompose the code into domains 3/ Re-architect the code into microservices in cloud 4/ Transform mainframe code into modern language 5/ Test to ensure functional equivalency 6/ Deploy to the cloud. Those steps used to be completed through siloed workflow and complex documentation. With Mainframe Modernization service, user has one-stop solution to follow through all the steps with reduced friction, resulting in faster and smoother goal achievement.

Goal

Unify different workflows and tools of mainframe modernization into a service hub.

My Role

• Lead the customer journey to identify workflows.
• Design easy to use dashboard and workflows to guide user through modernization journey.
• Design ways to map mainframe data into modern user interface.
• Work with PM and dev to implement design and ensure launch quality.

Persona

There are variety of persona involved in the mainframe modernization process, to name a few:

• Application owner
• Migration engineer
• Modernization consultant
• Project manager

Customer journey

Customer's journey are mapped into swim lane labeled with goals, tasks, deliverables,personas for AWS/partner/customer, tools, challenges and opportunities. Journey map provides team with holistic overview of the process to highlight painpoints and challenge. Ideas spring from journey map, and user stories were made afterwards.

Due to sensitivity of the information, only an outline and blurred version of this artifact is shown.

Results

Follwo up

This project later becomes the foundation of AWS Transform, a platform to modernize all types of legacy code and data centers, such as VMware, .NET. Lots of the data in this project were indexed through LLM so user can interact with chat agent to get information. The data here were used as evidence to check if chat response were accurate. We also did extensive usability study regarding whether chat could replace tabular display of large volumes of data, the responses have always been that they want both, on the grounds of 1/ Chat was slow returning data queries 2/ Users still need a persistent display of the data to understand the relationships. As a result in the later development, both chat and full volume data presentation were used for the product.