AI Backend Tooling
In designing the error notification system within Power Query, I aimed to create a user-friendly experience that enhances productivity and minimizes frustration for users of Dynamics 365 and Microsoft Word. Power Query serves as a vital conduit for users to import their personal data into the Azure Data Lake, making accurate data transformation crucial. My approach began with analyzing common user errors, prioritizing frequent issues encountered during this critical data process. I implemented a layered notification system, featuring real-time alerts that are both informative and actionable. By leveraging an AI-enhanced engine, the system offers contextual suggestions for resolving errors, streamlining troubleshooting. I also ensured that notifications can be customized, allowing users to set preferences for alert types and frequencies based on their workflows. User testing was essential; feedback led to continuous improvements in the clarity and relevance of notifications. This system empowers users with faster problem resolution and complements the sophisticated capabilities of Dynamics 365 and Microsoft Word, enhancing overall user satisfaction.
Features a dedication panel for error handling
USE CASE – DEEP DIVE
One centralized location to work through fixing errors in an inefficient way.
Defined 2 distinct error types.
With 3 unique UX interaction workflows.
Broken errors are the most intrusive of the error types. These must be fixed before the system can leverage any of the intended data.
Invalid errors of the Challenge type are also data compromised only because we must review the proper credentials to proceed.
Invalid errors of the Validation type are the least sever and can be addressed at any point. They do not limit the user from consuming the data.
We do not limit the user from saving his query mid-workstream.
We know our users and we understand that fixing errors can be time-consuming, We do not want to ever limit the user from being about to walk way and come back at another time to continue his/her query work.
