Foundations
the data-science bedrock every model stands on
3 pagesThe Road to LLMsHow seven decades of AI led to models that understand and generate language.9 deep dives · 4 key termsData Science & ML FundamentalsThe discipline of turning data into predictions - the bedrock under all AI.8 deep dives · 4 key termsNeural Networks & the TransformerThe architecture family - and the one design - that made LLMs possible.7 deep dives · 4 key terms
Models
the engine you call — and how to change what it does
3 pagesLLMs & Foundation ModelsLarge models pre-trained on vast data, adaptable to almost any task.9 deep dives · 4 key termsEmbeddings & Vector SearchTurning text and images into meaning-carrying numbers you can search.33 deep dives · 4 key termsFine-Tuning & AlignmentActually changing the model's weights for your task, style, or values.20 deep dives · 4 key terms
Ground
making a general model useful for your job
3 pagesPrompt EngineeringSteering the model with how you ask - no training required.22 deep dives · 4 key termsRAG - Retrieval-Augmented GenerationGive the model your data at answer-time instead of retraining it.17 deep dives · 4 key termsContext EngineeringDesigning what actually occupies the context window - budget, order, and what to leave out.
Build
wiring models into real, multi-step products
5 pagesOrchestration FrameworksThe glue that chains models, data, and tools into a real workflow.10 deep dives · 4 key termsAgents & Tool UseModels that plan, call tools, and take multi-step actions toward a goal.27 deep dives · 4 key termsAgent SkillsWhat a skill actually is, which surface it runs on, and what it is constantly confused with — every claim dated, because provider surfaces move monthly.35 deep dives · 5 key termsData for AIThe pipelines, ownership and quality work that decide whether any of the rest can work.11 deep dives · 13 sectionsForward-Deployed EngineeringEngineering that starts at the customer's workflow and ends at a measured outcome — coding is only part of the job.4 deep dives · 4 key terms
Operate
making it trustworthy, measurable, affordable and production-ready
8 pagesEvals & TestingHow you know the system is actually good - and staying good.11 deep dives · 4 key termsGuardrails & Responsible AIKeeping outputs safe, accurate, private, and compliant.14 deep dives · 4 key termsDeployment, Inference & LLMOpsRunning models reliably and affordably at scale - MLOps for LLMs.26 deep dives · 4 key termsWhat Happens After You Hit EnterThe forty-two stages between your API call and the first character on screen — one fixed sequence, not a menu, and any one of them can be why the answer is wrong or slow.42 deep dives · 5 key termsAI CostingHow AI cost is determined and controlled — token math, the model-selection matrix, unit economics, and the levers that move the bill.10 sectionsFailure ModesHow AI systems break in production, what each failure looks like, and what fixes it.10 sectionsSecurityPrompt injection, data exfiltration, and the tool permissions an agent should never hold.8 deep dives · 13 sectionsEnterprise AI OperationsOne delivery system seen from three seats — what developers, architects and project managers are each judged on, and what a bad number actually means.4 sections
Reference
consult, don't read — look it up and run the numbers
9 pagesThe ToolingFor every AI element, the tools that matter and why each is chosen — each one linked to the page that teaches what it does.2 sectionsFrontier ModelsCurrent frontier LLMs — context windows, cost per million tokens, and what each is for. A leader's snapshot.3 sectionsThe MapEvery topic on one page — one-liner, key types, a key term and the question you're most likely asked.7 sectionsOpen-source resourcesThe repositories people actually build LLM products on, and the situation each one is for.9 sectionsChecked, and wrongClaims about AI models that do not survive a primary-source check — each one dated, and carrying the method so you can re-run it.8 sectionsROI DashboardPortfolio ROI with drill-down — put your own numbers in and see which AI bets pay back.4 sectionsBuild vs BuyTool vs in-house, element by element — the scoring behind the call.4 sectionsEvals DashboardA worked LLM-as-judge error analysis you can read end to end — evals in action, not in theory.6 sectionsThanks & what I learnedThe courses I learned from, and where that shows up on this site.4 sections