Microsoft AI MVP designs a Microsoft Teams bot to centralize tools for employees

Ayca Bas

February 22nd, 20220 0

Learn from the community is a series of articles to highlight real-life project scenarios covered in Show & Tell interviews at Microsoft 365 Developer YouTube. It’s a great opportunity for me to inspire you with the end-to-end solutions built upon Microsoft Cloud and help you for your next project with useful resources.

A lot of us use many tools in our day-to-day work and these tools vary according to our job responsibilities. Microsoft AI MVP Stephan Bisser joins me in this interview to talk about how tools vary between departments to get a job done in a company. For example, when the Sales department uses CRM tool to track new customers, developers in the same company often use Azure DevOps to track their software development process on a day-to-day basis. Stephan explains how hopping between the internal tools became a challenge and how they decided to build Erdwin bot to solve the problem. Erdwin bot is designed to centralize tools across the company and help employees access the data they need through one single bot interface. There is something for every employee in this bot.

Erdwin bot is built using the Microsoft Bot Framework SDK and it is designed as a Microsoft Teams bot. It also leverages Microsoft Cognitive Services including:

Language Understanding (LUIS) to enable intelligent understanding with natural language processing
QnAMaker to create knowledgebases to handle user’s basic questions.

Erdwin bot connects to many tools that are used internally in Stephan’s company. It provides the following features and capabilities:

Managing and gathering opportunities on Dynamics 365 CRM by using Dynamics 365 Customer Engagement Rest APIs
Handling task management on Microsoft Planner by using Microsoft Graph APIs
Alerting employees about important news by connecting to the Intranet
Helping employees’ onboarding with Q&A capability as single place of truth by using QnAMaker API
Getting backlog items by using DevOps APIs