Katherine Johnson (AI Generated Artwork)
Machine Learning Fundamentals at Repo Racers
This guideline outlines the Machine Learning (ML) practices at Repo Racers. We collaborate with clients to develop ML models and deploy them into production environments, focusing on adhering to engineering and research best practices throughout the project lifecycle.
Goals
- Establish a comprehensive set of ML practices for project execution.
- Clarify the ML process and its integration within broader software engineering projects.
- Offer best practices for various stages of an ML project.
How to Use These Fundamentals
- Starting a new ML project? Begin with our general guidance documents.
- Looking for advice on specific parts of your ML project? Consult the guidelines for the different project phases.
ML Project Phases
Below is a diagram illustrating the phases of an ideal ML project. While practical limitations and specific requirements might prevent a project from strictly adhering to this structure, best practices for each phase should still be followed.
- Envisioning: Initial understanding of the problem, along with customer goals and objectives.
- Feasibility Study: Evaluation of whether the problem can be effectively solved with ML given the available data.
- Model Milestone: Achievement of a baseline model meeting minimum performance criteria, both in ML metrics and system performance. Use insights gained up to this point to define the project's scope, objectives, high-level architecture, definition of done, and overall plan.
- Model(s) Experimentation: Tools and best practices for conducting effective model experimentation.
- Model(s) Operationalization: Checklist for model readiness for production.
General Guidance
- ML Process Guidance
- ML Fundamentals Checklist
- Data Exploration
- Agile ML Development
- Testing Data Science and ML Ops Code
- Profiling Machine Learning and ML Ops Code
- Responsible AI
- Program Management for ML Projects