Practicum 3: Team Industrial Project with SLA and Quality Constraints
A team industrial project: production ML with real SLA, latency, and quality constraints
Instructor to be announced
About the course
The third-semester practicum is the program's flagship engineering project. Teams of 3-4 work on a challenge from an industry partner or a real open-source project with production-level requirements: SLA, latency budget, and monitored quality. Students apply skills from every previous course under real-world constraints. The outcome is a delivered service complete with an SLO report, a postmortem, and a public demo.
What you'll learn
✓Deliver a team ML project with explicit SLOs and documented tradeoff decisions
✓Run load testing and optimize a system to meet a latency budget
✓Write a production postmortem with quantitative root-cause analysis
Key topics
•Requirements decomposition and ML system design under constraints
•Architectural decisions: latency vs. accuracy tradeoffs
•Agile process in an ML team: sprints, Definition of Done for ML
•Load testing and capacity planning
•SLA/SLO/SLI: design and measurement
•Production postmortem and root-cause analysis (RCA)
•Demo and technical presentation for a non-technical audience
This description was generated automatically and has not yet been reviewed by an instructor — it's a draft for discussion.