Master prompt, context, harness, loop, and graph engineering alongside current agent architectures, evaluation methods, and fast-moving AI technology.
Learn AI insights
The model is only one component of a capable AI system. Exalted AI connects prompts, context, tools, state, harnesses, feedback loops, and graph architectures so you can understand how modern agents are actually engineered.
Experience behind the lessons
Joe Chang brings more than 20 years of software engineering experience across startups, research, international banking and FinTech, entrepreneurship, and production systems. He has taught more than 1,000 students across machine learning, deep learning, reinforcement learning, large language models, and AI engineering. Exalted AI extends that teaching into an advanced technical reference for building modern agentic systems.
Engineering tracks
Move across prompt, context, harness, loop, graph, Python, and current AI technology topics according to the system layer you need to understand.
Use advanced Python mechanics—typing, callables, generators, structured state, async, concurrency, and error boundaries—to build reliable AI systems.
Go learning pathEngineer model behavior as an interface through instructions, examples, reasoning scaffolds, output contracts, chaining, and systematic evaluation.
Go learning pathControl what the model knows at each decision through retrieval, memory, routing, compression, isolation, and long-horizon context management.
Go learning pathBuild the software layer around the model with tools, skills, MCP, state, execution, authorization, recovery, persistence, and observability.
Go learning pathDesign bounded feedback loops that act, observe, evaluate, revise, preserve useful state, and stop under explicit conditions and budgets.
Go learning pathCoordinate agents, state, dependencies, and control through graph-based architectures that connect larger AI systems and multi-agent workflows.
Go learning pathUnderstand important new AI systems and ideas through mechanism-first explainers on projects, protocols, research, and architecture debates.
Go learning pathFeatured tutorials
These articles focus on architecture, state, context, feedback, evaluation, and practical consequences rather than launch-demo vocabulary.
An agent that has run forty steps and produced an answer no better than step three is not stuck. It is broken in a specific, nameable way.
Read tutorial
“Everything is a plugin” is a claim about architecture, not a feature list. It says the agent loop itself is replaceable — and that is expensive to mean.
Read tutorial
Two teams ship the same model, the same prompt, the same tool list. One agent finishes a multi-step task. The other stalls, loops, and quietly corrupts its…
Read tutorial
The prompt you spent a week tuning is not what the model reads. It reads everything in the window — and everything else is quietly voting.
Read tutorial
The demo works. Then you add a second worker, a shared artifact store, and a verifier, and the system starts producing results nobody can attribute,…
Read tutorial
A prompt is a spark, not an engine. The spark still matters — it just cannot carry the machine.
Read tutorialLatest
Fresh technical explanations of LLM systems, agents, tooling, evaluation, and emerging engineering patterns.
MCP answers "what can I call?" A2A answers "who can I hand this to?" Different questions, different failure modes.
Read tutorial
A dict payload that passes every test until someone renames a key. A callback whose signature drifts by one argument. A duck-typed object that satisfies…
Read tutorial
Your recall metric looks healthy. Your answers are still wrong. That gap is the whole reason advanced retrieval patterns exist.
Read tutorial
The demo worked because the agent was running on your laptop. It had your credentials, your network, your home directory, and your shell history. Move it…
Read tutorial
The demo agent works. It reads the file, runs the test, patches the bug. Then it ships to a second tenant, the session runs for six hours, the process dies…
Read tutorial
A prompt fix that passes the demo and fails two weeks later is not a bad prompt. It is an improvement process with no memory.
Read tutorialBuild a complete mental model of modern agent engineering. The Agent Engineering Starter Pack connects prompt interfaces, context, memory, workflow design, agent harnesses, tools, persistent state, feedback loops, evaluation, and graph architecture—so you can understand how capable agent systems are assembled layer by layer.
You’ll receive the bundle by email. You can unsubscribe anytime.
