Software Systems: From Code to Production

Learn to connect an algorithm's correctness argument to the comparisons, branches, and memory accesses it performs.

Compare arrays, hash tables, trees, heaps, and graphs by the operations they make cheap and the memory accesses they require.

Follow compilation, linking, loading, address spaces, and the first instructions that turn a program into a running process.

Understand how processes, virtual memory, scheduling, files, and system calls let programs share a machine.

Follow an instruction through a pipelined processor and explain why cache locality often matters more than operation count.

Follow an HTTPS API request from a browser through name resolution, transport, TLS, HTTP, server code, and back.

Trace SQL through parsing, logical rewrites, cost-based planning, physical operators, indexes, and the buffer pool.

Follow a small program through lexing, parsing, semantic checks, intermediate forms, code generation, and runtime execution.

Separate JavaScript callbacks, event-loop scheduling, operating-system I/O, and worker-pool tasks in an asynchronous program.

Design timeouts, retries, idempotency keys, and server-side state so a lost response does not silently duplicate an operation.

Use cohesion, coupling, and dependency direction to choose software boundaries that make likely changes easier to contain.

Turn user requirements into a data flow, capacity model, and failure-aware architecture before drawing infrastructure boxes.

Understand why timeouts, retries, ordering, and consistency matter when remote machines can fail independently.

Follow the cloud control plane that compares declared configuration with live resources and works toward the requested state.

Threat-model a request by tracing sensitive data, user authority, trust boundaries, and the controls that limit abuse.

Use workload shape, latency distributions, profiling, and controlled experiments to find the cause of slow software.

Follow a change from Git objects and review through tests, build artifacts, deployment, rollback, and production feedback.

Walk one fresh HTTP/2 request over TCP from URL parsing through DNS, TLS, server handling, and the response.

Follow a selective SQL lookup through planning, a B-tree index, heap visibility checks, and the result rows.

Follow a client command from the leader's log through replication, quorum commitment, state-machine application, and reply.

Visualize review, automated checks, immutable artifacts, staged rollout, observability, and rollback decisions.