10 Post project (sprint Ω)

 

This chapter covers:

  • Looking after an ML system after it’s gone into production
  • Dealing with production failures
  • Learning from the project and improving practice

The models are integrated in an application, and the application is delivered to production. Now someone must look after it! In addition to dealing with old ML systems and looking after new ones, this chapter addresses what happens to the team after you complete the project. How can you and your team learn from it, and what should be done to make the next project better?

10.1 Sprint Ω backlog

The backlog in table 10.1 lays out the work that the team needs to cover once you’ve delivered a system into production.

Table 10.1 Sprint Ω backlog (view table figure)

Task #

Item

SΩ.1

Identify ML-specific sources of technical debt.

  ▪ Verify model performance as appropriate

  ▪ Monitor model drift

  ▪ Check model obsolescence

SΩ.2

Identify and deal with technical debt in general.

SΩ.3

Run a post-project review to determine what the team can learn from your project.

SΩ.4

Seek ways to develop new practices for developing ML systems.

SΩ.5

Identify new technologies that the team can use to be more successful.

SΩ.6

Write a case study about the project to record and share your experience.

10.2 Off your hands and into production?

10.2.1 Getting a grip

10.2.2 ML technical debt and model drift

10.2.3 Retraining

10.2.4 In an emergency

10.2.5 Problems in review

10.3 Team post-project review

10.4 Improving practice

10.5 New technology adoption

sitemap