Topics
Lectures from every course, grouped by subject — so ideas that show up in different courses sit side by side.
- Distributed systems 6
- Architecture 10
- Networking & proxies 4
- Containers 4
- Orchestration 1
- Agentic software engineering 6
- Language models & inference 2
- Retrieval & RAG 1
- Tools & MCP 2
- Evaluation & reproducibility 7
- Agent security 2
Distributed systems
6 lecturesArchitecture
10 lectures- Agent topology, context and memory as recorded decisions
- Specs, locked tests and layer rules as delivery gates
- Harness architectures compared on cost, time and success
- Reproduced results, validity threats and thesis paths
- Microservices, HTTP APIs and Docker
- Docker Compose, NGINX and horizontal scaling
- Concurrency, idempotency and Saga workflows
- Coordination, failure detection and leader election
- Replication, consistency, recovery and observability
- Introduction
Networking & proxies
4 lecturesContainers
4 lecturesOrchestration
1 lectureAgentic software engineering
6 lectures- Agent anatomy, reliability and the token cost model
- The agent loop as a tested, evaluated component
- Agent topology, context and memory as recorded decisions
- Silent harness faults and configuration fitness checks
- Specs, locked tests and layer rules as delivery gates
- Calibrated graders and trajectories as evaluation evidence
Language models & inference
2 lecturesRetrieval & RAG
1 lectureTools & MCP
2 lecturesEvaluation & reproducibility
7 lectures- Inference capacity and structured output as a contract
- The agent loop as a tested, evaluated component
- Retrieval as a tested subsystem
- Silent harness faults and configuration fitness checks
- Calibrated graders and trajectories as evaluation evidence
- Harness architectures compared on cost, time and success
- Reproduced results, validity threats and thesis paths