Elements of Scalable Web Development

Aalto University course book

Elements of Scalable Web Development

Architecture, measurement, scaling pressure, and agentic workflows.

Start reading

Try it in the book

Predict. Experiment. Explain.

Explore demos and check your understanding throughout the chapters. Switch between Explore and Practice, check your answers, and sign in to save your results for another visit.

Start with measurement

Course project

One contract. Three fresh builds.

Build the Work Activity Tracker with a large model, a small model, and an agent fleet. The behavior stays stable; the development system changes.

Start the course project

Book contents

Front matter

  1. Introduction Elements of Scalable Web Development

Part I: Establish — Build and measure a baseline

  1. Chapter 1 What does scalability mean for web development?
  2. Chapter 2 How can coding agents support scalable development?
  3. Chapter 3 How can we measure web applications and development workflows?

Part II: Challenge — Expose architectural assumptions

  1. Chapter 4 How does web architecture shape scalability?
  2. Chapter 5 How does development friction limit web projects?
  3. Chapter 6 Which web application assumptions fail under pressure?

Part III: Compare — Evaluate workflows and scaling responses

  1. Chapter 7 When do agent fleets improve web development?
  2. Chapter 8 How should web applications respond to scaling pressure?
  3. Chapter 9 What do three agent-assisted implementations teach us about web scalability?

Appendices

  1. Appendix A Evaluating agentic development workflows
  2. Appendix B Reproducible development environments
  3. Appendix C Advanced control boundaries in web applications
  4. Appendix D Agent-system security, control, and traceability
  5. Appendix E Formal scalability and reliability models
  6. Appendix F Durable workflows and retry-safe effects

Field evidence

Case studies

Project histories that connect the book’s concepts to concrete development decisions, failures, and lessons.

Browse the collection
  1. Slideotter – Learning faster than the product converged How checks on rendered slides made the artifact inspectable while product breadth and delivery choices outran evidence of repeated use.
  2. Thesis Journey Tracker – Small scope made production success possible How a narrow product boundary and simple runtime turned rapid agent-assisted development into a working production tool.
  3. Kirjolab – When code production scaled faster than architecture How behavioral checks kept rapid development repairable while delayed feedback about ownership and dependencies made architectural consolidation expensive.
  4. Webernetes – Comparing generated behavior with a real cluster How an agent-assisted port used reviewable source mappings and shared tests against a real cluster to make selected compatibility claims observable.
  5. Modern Web Guidance – Treating current knowledge as a tested dependency How fast-changing web-platform knowledge can become a retrieved, versioned, and task-evaluated development dependency.
  6. Temper – Contracts, verification, and bounded autonomy How Datadog connects system verification, agent-authored behavioral contracts, and bounded live optimization to explicit acceptance evidence.