formerly AI Applications in Computer Science
AI Engineering is a project-first course about building dependable software with models, agents, data, and real interfaces—not simply learning how to chat with an AI.
The 2026 curriculum moves from command-line and data foundations into LLMs, context, APIs, databases, agents, deployment, and a shared Habitat system operating in the live Kepler world.
A shared world for engineering
Kepler is a simulated planetary environment. Habitat is the software students build to manage resources, crew, and construction in that world. Their systems have to work together in a setting that keeps changing.
Try a sample learning activity →What students learn
Engineer with AI
Use models and coding agents as capable tools while retaining architectural judgment and verification.
Design the context
Shape prompts, files, data, interfaces, and instructions so an agent can do reliable work.
Ship a system
Connect software, servers, persistence, live events, and human operations into one working whole.
Curriculum arc
Computer science and digital foundations
A systems-level introduction to computing, AI, and the path from physical signals and stored data to software that can reason over language.
Developer setup, terminals, and command-line tools
Installing the working toolchain, navigating the terminal, understanding command structure, and learning to discover an unfamiliar CLI rather than treating it as a black box.
Markdown, JSON, TypeScript/Bun, Git, and GitHub
Representing documents and structured data, building typed command-line software, managing repositories, and using version control as part of everyday engineering work.
LLM fundamentals: models, tokens, and chat training
Base models, tokenization, chat training, model behavior, and the technical distinctions behind systems that can generate and respond to language.
Prompt engineering, context engineering, and agent skills
Designing prompts, AGENTS.md files, reusable skills, and working context so coding agents can operate reliably within a real project.
Software design, object-oriented programming, and CLI development
Turning requirements into objects, commands, and discoverable interfaces while building the Habitat CLI for both human and agent users.
Machine learning, reinforcement learning, embeddings, and semantic search
Connecting learning concepts to practical systems through reinforcement learning, text embeddings, databases, and search based on meaning rather than exact keywords.
APIs, HTTP, REST, DNS, authentication, and cryptographic keys
Working across network boundaries with requests, RESTful interfaces, API tokens, environment files, DNS, asymmetric keys, and secure service access.
Servers, deployment, systemd, OpenClaw, MCP, and WebSockets
Deploying software to a real server, keeping it running, connecting agent infrastructure, and adding live event communication between services and interfaces.
Kepler/Habitat capstone development
Teams integrate blueprints, construction, crew, exploration, collection, alerts, dashboards, live Kepler ticks, and agent capabilities into one working system and final presentation.
Projects
The 2026 course uses one evolving system as the laboratory. Each new idea has to survive contact with the code, the server, other teams, and a world that keeps moving.
A capable Habitat CLI
A discoverable command surface for humans and agents, backed by a growing domain model.
A deployed service
Persistent data, REST interfaces, secure remote infrastructure, live processes, and operational recovery.
A connected world
A dashboard and agent-enabled system responding to shared Kepler events in real time.
Start a conversation
Tell me about your process, your students, or the software you want to build. We begin with goals, constraints, and timing, then agree on scope, deliverables, and fees before work begins.
Email Kevin ↗