MICROSOFT

Designing ES Chat

Skills Used

Natural Language Processing

Process Management

Layout Design

UX/UI Design

UX Research

THE PROBLEM

How can we simplify the task management process for non developers?

With all its features, Azure DevOps can be confusing to non developers, but its very common for Project Managers to use the tool to help plan and manage software development projects. From there, Project Managers relay the information to other members in their team on other platforms. So how can we make this process just a little more seamless for product managers and developers?

RESEARCH

Design constraints and considerations

When speaking with upper management, Product Managers, and Developers about the project, there were a few things that I wanted to keep in mind while working on this assignment.

Don't alienate current users

ADO has a solid userbase and we don’t want to create any dramatic changes that would upset current usability or workflows.

ADO might change in the future

With Microsoft constantly innovating and acquiring more development tools there is a chance that ADO could be a completely different program in the future and its something we should keep in mind.

This tool could become public

This current solve is to improve the workflow between product managers and software developers at Microsoft but there is a chance this could be a feature that everyone using ADO could use.

User Research

How does ADO fit into your workflow?

Reached out to Microsoft project managers and developers to get an idea of how they feel about the current ADO experience. With ADO being such a large project, I wanted to get a good idea of what pain points exist and maybe find a commonality amongst them.

THE PROBLEM

Looking into Microsoft Teams and ADO integrations

Looking deeper into Microsoft’s apps, there are app integrations directly related to ADO that developers and project managers are already using.

PROPOSED SOLUTION

What about a chatbot?

At their core, chatbots interpret information given to them, find an answer, and return it to the user in a natural language format. This reminded me a lot of a secretary or an interpreter.

Benefit #1: Communicate between the already existing Microsoft Teams integrations

ADO has a solid userbase and we don’t want to create any dramatic changes that would upset current usability or workflows.

Benefit #2: Ability to look through previous team messages and relay info

With Microsoft constantly innovating and acquiring more development tools there is a chance that ADO could be a completely different program in the future and its something we should keep in mind.

Benefit #3: Integration between ADO and Microsoft Teams

Users would be able to continue conversations and messages across platforms

EXECUTION

But... Microsoft already has a ton of Chatbots

Microsoft Teams already has an AI chatbot and with Microsoft acquiring GitHub (a software very similar to ADO) there is also an AI chatbot coming from there. The goal is not to create a new chatbot but creating a secretary for all of the chatbots.

MONITORING FEEDBACK

Feedback in ADO and Teams

An important part of expanding this project and seeing if its actually feasible, is to release a simplistic version for internal use and be able and use feedback in order to make improvements in the future.

Feedback can be obtained via Azure Dev Ops and Microsoft Teams in a few ways. We're asking users for simple thumbs up thumbs down, and then prompting if they want to give us more information from there.

SOLUTION

ES Chat in Microsoft Teams

ES Chat works as a secretary between the user and Microsoft’s many chatbots. In teams, the focus is on helping users to summarize documents and meetings between chats and create documents. When a user asks ES Chat a question, ES Chat checks in with other chatbots (Copilot, Bing, etc.) to come back with answers.

Select gallery images to increase size.

SOLUTION

ES Chat in Azure Dev Ops

Within Azure Dev Ops (ADO), ES Chat works through GitHub Copilot, primarily pulling information and answers from the ai chatbot as well as Bing.

An important aspect of ES Chat is feedback. Within ADO, feed back is going to be one of the main drivers for how users will interact with the chatbot in the future.