The complete craft of building with AI

The Builder's Stack.

A working education for people who build with AI: the technology underneath the tools, the skill of working with models, and the human factors that decide whether people trust what you ship. Practical, current, and written to be used.

Part I
The Software Map
The big picture of how software actually works.
0/7
01Decode the questionsNext
02What software is
03Where software lives
04How a request travels
05Languages & frameworks
06Git, your undo button
07Environments
Part II
The Technology Stack
The layers every product is built from.
0/7
01Frontend
02Backend
03APIs
04Data
05Hosting
06Auth
07Monitoring
Part III
The Model Layer
Scope your build and wire in the model, one rung at a time.
0/7
01Find your build
02Model vs. coding assistant
03Make your first call
04Prompt like an engineer
05Output your code can use
06Give it the right context
07Give it tools
Part IV
Ship and Track
Choose the model you ship, run the pre-flight, and learn from real use.
0/6
01Choose a model
02Wire in the model
03Review the build
04The pre-flight
05The Build Plan
06After the ship
Open The Fundamentals
Part I
Human Factors and Design Principles
Experience design principles for AI products, grounded in the cognitive science.
0/8
01Why human factorsNext
02Perception: get noticed
03Working memory: keep context
04Anxiety: lower the stakes
05Mental models: what it can do
06Metacognition: catch errors
07Supervision: human in charge
08The human factors audit
Part II
Evals
Measure whether your AI is actually any good.
0/8
01The demo isn't the product
02The quality bar: what good is
03Test cases: sample reality
04Graders: score the outputs
05The regression gate
06Production signals
07Goodhart's trap: gamed metrics
08Build your eval
Part III
Agentic systems
When your product starts doing things, not just saying them.
0/8
01When AI takes action
02The action surface
03The autonomy ladder
04Blast radius: limit the damage
05Prompt injection
06Receipts and recovery
07Agent evals: judge the run
08The Agent Charter
Part IV
Security
Defend your product before attackers test it for you.
0/8
01Why AI gets attacked
02Threat modeling
03Identity: whose keys
04Data: what leaks out
05The supply chain
06Defense in layers
07Red-team your product
08The Security Posture
Open The Practice
Part I
Orchestration
Running agent fleets in parallel.
0/8
01Beyond one agentNext
02Decomposition: split work
03Patterns: fan-outs, pipelines
04Fleet verification
05Economics: what fleets cost
06Fleet failure modes
07Supervision at scale
08The Orchestration Plan
Part II
AI in high-stakes industries
Shipping AI into finance, healthcare, and law.
0/8
01Why high-stakes differs
02Boundaries: what AI can't do
03Choosing the model
04The right context
05Security and guardrails
06Proving fairness
07Testing & evidence
08The High-Stakes Clearance
Part III
Context & Memory
Deciding what your product knows and remembers.
0/8
01Context failures
02Retrieval: look it up
03The search stack
04Memory: what it keeps
05Context budgets
06Freshness and conflicts
07Grounding evals
08The Knowledge Charter
Part IV
The Unit Economics of AI
Making every answer cost less than it is worth.
0/8
01The cost cliff
02Cost per task
03Latency: design the wait
04Routing and cascades
05Caching: pay once
06Batch and background
07Pricing: match your costs
08The Inference Budget
Part V
When Your Users Are Agents
Designing your product for the agents that act on behalf of people.
0/8
01The user without eyes
02The tool storefront
03Discovery: get found
04The delegation chain
05Guardrails for guests
06Computer-use traffic
07When the buyer is a bot
08The Agent Access Policy
Part VI
Running an AI Organization
Managing AI product operations end to end.
0/8
01The operating layer
02Ownership: one name
03Change control: release it
04Govern the knowledge
05Govern access and safety
06Incident response
07Build the team
08The Operations Charter
Open The Frontier

The Fundamentals

The Practice

The Frontier