Speakeasy provides an essential service that enables companies to generate SDKs through open API specifications. Operating as a critical middleware, Speakeasy facilitates a vital connection between companies and their SDKs, making a smooth and efficient customer experience paramount.
Encountering the Challenge
A significant hurdle for Speakeasy was managing and keeping track of the open issues on GitHub faced by their customers. To address this efficiently, we crafted an automated workflow using Retool, which not only monitors these issues but also sends timely notifications to Speakeasy's internal team, ensuring proactive customer support.
Rapid Development and Execution
What might typically take several weeks to develop, our Retool developers achieved in a mere 25 hours. This comprehensive solution covered everything from data retrieval, backend and frontend workflows, to the UI design on Retool.
Workflow Automation for Customer Support Designed by Toolpioneers' Retool Developers
The Daily Workflow
Retool workflow initiates a daily cron job.
This job fetches repository details from GitHub using Graphql query.
Open issues from each repository are then retrieved.
Pertinent details about these issues - including the title, URL, description, and creation date - are extracted using a REST API.
These details are updated onto a table within a Postgres database. This table is refreshed daily.
For easy tagging of the responsibility of the issue resolution, the table contains a column named 'responsibility'. By default, this is set to 'speakeasy'.
If any issue is closed on GitHub, its corresponding record is deleted from the Postgres database, ensuring the database remains up-to-date and free of resolved issues
The Notification Workflow
Hourly Notifications: Every hour, the system scans the repositories for any new issues that have emerged in the past 60 minutes. These are then formatted and sent as notifications to a dedicated Slack channel for real-time monitoring.
Weekly Notifications: Once a week, the system reviews the Postgres database for issues where: - Responsibility is assigned to 'speakeasy'. - The issue remains open. - The issue was created within the last seven days. All qualifying issues are compiled and sent as a bulk Slack notification.
Both the workflows ensure that no issue slips through
Assigning Responsibility
The accuracy of issue identification is crucial. Speakeasy has to determine whether they are responsible for an issue or if it lies with the customer.
For this, an 'Issue status' dashboard was designed. While the system defaults to assigning responsibility to 'speakeasy', the team has the flexibility to reassign responsibility to the customer if deemed appropriate.
Once this change is made, an 'Upsert' button ensures any changes made within the app are accurately reflected in the main Postgres database. As a result, issues marked as the customer's responsibility are excluded from the weekly Slack notifications
Conclusion
The initiative by our Retool developers at Toolpioneers to enhance Speakeasy's customer support through workflow automation goes beyond mere problem tracking; it's about ensuring issues are addressed promptly and efficiently. Through meticulous data management, real-time alerts, and seamless app integration, we've set a new standard for proactive issue resolution in customer support. Explore our workflow here