Track · 21 lessons · 10.3 h
AI Systems & MLOps
Serve, scale and operate models and AI services in production.
0 / 21 done
MLOps foundations4 lessons · 1h 50m
- What Is MLOps and Why AI Engineers Need Iteasy15 min
- Drift Detection: Data Drift vs Concept Drift (KS, PSI)medium30 min
- Experiment Tracking with MLflowmedium35 min
- MLflow Model-Stage Promotion: Staging → Productionmedium30 min
Serving & APIs5 lessons · 2h 40m
- Serve Models with FastAPI + a Model Registryhard50 min
- FastAPI Architecture for AI Servicesmedium35 min
- Designing API Contracts for AI Servicesmedium30 min
- Streaming Responses with SSE & WebSocketseasy15 min
- Session Memory Stores: Redis-Backed Conversationsmedium30 min
Performance & scale5 lessons · 2h 15m
- Semantic Caching to Cut Latency & Costeasy15 min
- Model Quantization: FP32 → INT8 & Micro-Batchingmedium30 min
- Async & Parallel Pipelines: Hiding Latencymedium30 min
- Horizontally Scaling AI Servicesmedium30 min
- Load-Testing AI Services with Locustmedium30 min
Reliability & operations5 lessons · 2h 35m
- Model Fallback & Circuit Breakers for HAmedium30 min
- Structured Logging Middleware for AI Servicesmedium30 min
- Cost Tracking & Telemetry for AI Appsmedium30 min
- Document Ingestion Pipelines (Connectors, OCR, PII)medium35 min
- Async Job Tracking for Long-Running Ingestionmedium30 min
Interview prep2 lessons · 55 min
- How to Explain MLOps & Drift in an Interviewmedium25 min
- System Design: Scalable Inference Servicehard30 min