Clarifai is an artificial intelligence company that provides a visual recognition API for app developers to create and train their own machine learning models. A user uploads an image or video, and Clarifai’s models return tags about what the image contains along with applicable probabilities, powering a more intelligent visual search.
Founded in 2013 in New York, Clarifai uses machine learning and deep neural networks to provide a powerful image recognition system used across industries. The technology team has grown to 25 engineers focused on data science, research, front end and back end development. Clarifai has 9 machine learning models written primarily in Python and Go.
Postman for development and testing
Every engineer at Clarifai is responsible for testing their own code and completing code reviews for their teammates as well. When a developer at Clarifai has an idea for a new endpoint, they create the endpoint in Postman first. They write tests in Postman to ensure their responses are returning as expected and establish a proof of concept.
Then the engineer shares the Postman Collection or folder with the rest of the team to test it against their own scripts. At every step of the release cycle, Clarifai uses shared Postman environments to effectively test in development, staging, and finally production.
Collaboration and team communication
Clarifai’s primary Postman Collection contains 179 endpoints. The engineers review changes to the endpoints by tracking the activity feed in the Postman team library. It’s an easy way to stay up to date on what other teams are working on.
“Postman is the main tool we use internally to communicate any changes to our API. Every engineer at Clarifai uses it. Even people who aren’t engineers are using it. We have product and sales people who learned to code in order to test the Clarifai API in Postman, “ according to Jared Lerner, Senior Support Engineer at Clarifai.
If a client or prospect is experiencing a technical issue, the sales team can go into Postman to verify what the client is seeing. They’re able to reproduce the issue without having to rely on the engineering team for a diagnosis.
Postman monitors the health of their endpoints
Clarifai relies on Postman’s command-line tool, Newman, in conjunction with Jenkins to run automated tests for their continuous integration and continuous delivery processes.
Every 5 minutes, Clarifai uses Newman to run their Postman Collections and test every possible action on their public-facing endpoints. Those tests ensure that a healthy 200 status code returns from each API call, and the status is consequently reflected on the Clarifai website.
Customers rely on these status updates to know whether a service is down. Any Postman test failures will then trigger an alert notifying the Clarifai engineering team of the outages so they can get started on a solution immediately.
Postman is the main tool we use internally to communicate any changes to our API. Every engineer at Clarifai uses it.
The tool is robust, it’s easy to use, and it’s super efficient. I can get someone up to speed on Postman in seconds.
Jared Lerner, Senior Support Engineer