Please help with issues testing CrewAI Quickstart deployment on app.crewai.com

I just created a support request on app.crewai.com for this. I’m posting here on the community forum as well. Thanks for help!

Hi. I just upgraded to a CrewAI Basic Plan and tried to deploy a copy of the CrewAI quickstart at GitHub - royseto/crewai-quickstart: Copy of CrewAI quickstart from https://docs.crewai.com/quickstart as of Jan 2025. The CrewAI UI immediately says “Provisioning failed, try again.” I did add 3 values as environment variables on the launcher. It works for me locally.
Can you review and let me know whether I am doing some sort of pilot error?

3 EVs on launcher:

  • SERPER_API_KEY
  • OPENAI_API_KEY
  • MODEL=gpt-4o

Update: I got past the deploy error by committing my uv.lock file to GitHub and redeploying.

Now, my first test execution with topic “AI Agents” completed successfully in Execution Time 43.59s.

I am running a second test execution that I kicked off 15 minutes ago (around 1:30 PM PST on January 14, 2025). The first task completed in 18.003 secs but the second task is still "Pending Run 15 minutes later. This 15-minute delay to start running the second task seems unexpected.

How can this be resolved?

Update 2: this execution finally completed in 20 minutes. The second task was “Pending Run” for almost all of the 1218.714 secs (20 minutes) of its execution time as listed on the app.crewai.com console.

This worked functionally, but the 20-minute wait time to start running the second task was an unexpected performance issue. Are there best practices to make crew runtimes on app.crewai.com more consistent?