Friday, April 24, 2020

Optical Character Recognition - Event Attendance and Feedback Management



Business Case:

Our customer conducts an annual internal expo with numerous showcase events. Thousands of employees attend the events. It was a very time-consuming process to manually track attendance, send acknowledgement messages, get feedback, and compile insights from all the interactions.

Solution:

We created a PowerApps based solution that used Microsoft Cognitive Services for NLP and Image processing, Logic Apps for automation, Microsoft Teams for post event discussions and Power BI for reporting insights.

Solution Design:

We built a check-in app that uses the device’s built-in camera to capture the identification badge of each participant. The captured images are stored in an Azure Blob. An Azure logic app reads the content from the captured images utilizing Optical Character Recognition (OCR) API to check participation and update database attendance records. After the event, notifications are sent to attendees via Teams to fill out a feedback survey using a Flow Bot. The Feedback App captures and reports the survey responses to determine the CSAT (Customer Satisfaction) score of the event.

There are multiple checkpoints incorporated in the system to ensure authentication. All attendees’ Teams and mailing details are securely stored in a database for the period of the event. While authentication occurred through Azure AD, experience owners could use their mobile phones to access information through PowerApps.
Figure 1: Solution architecture

Application Preview:

App provides two separate interfaces for experience owners (employees) and event organizers (admins).

  Check-in app provides experience owners with a list of planned events and allows them to mark attendance. Feedback tab allows them to provide feedback for the events based on experience area and showcase details. 
  Admin app provides control over the events being organized. Users can manage attendance, and make necessary alterations as required by performing manual mapping for unidentified/inappropriate images of scanned badges. 




Figure 2: Application screenshots

Business Outcome:

The app increased the ease of capturing the event attendance. Our client used session attendance information to evaluate session adequacy and effectiveness. Also, in-depth analysis on end user feedback helped the client identify focus areas to improve user experience.

Highlights:

1.    Developed low code enterprise grade solution with minimal time to market
2.    Optimized the efforts in attendance tracking and feedback management of events
3.  Facilitated admins and organizers to understand users and improvise relevance and impact of program
4.    Reduced manual error
5.   Improved the efficiency of recording attendance and performing feedback assessment