Library
Search concepts, paths, labs, guides, and project tracks across the production AI stack.
Vector Spaces and Tensors
How coordinates shapes and transformations represent information numerically.
mappedUncertainty and Calibration
How probability estimates express uncertainty and align with outcomes.
mappedGradient Descent
How iterative optimization follows local objective information to update parameters.
mappedComplexity Analysis
How time and space growth guide algorithm and system choices.
mappedConcurrency Models
How asynchronous tasks threads processes and coordination affect correctness.
mappedSoftware Contracts
How interfaces invariants and tests preserve behavior across change.
mappedWeb Rendering and State
How server and browser rendering strategies shape interactive applications.
mappedReproducible Delivery
How automated builds tests images and releases create dependable deployments.
mappedDatabase Selection
How workload consistency access and scale determine storage choices.
mappedData Contracts and Lineage
How schemas quality checks and provenance make pipelines trustworthy.
mappedLearning Paradigm Selection
How available signals feedback and objectives determine a learning approach.
mappedGeneralization and Leakage
How evaluation design estimates future performance without contaminating evidence.
mappedBackpropagation
How gradients propagate through computational graphs to train neural networks.
mappedComputer Vision
Elective branch covering image representation recognition detection and generation.
mappedSpeech and Audio AI
Elective branch covering recognition synthesis understanding and audio generation.
mappedRecommendation Systems
Elective branch covering ranking personalization feedback and marketplace effects.
mappedTime-Series AI
Elective branch covering forecasting anomaly detection and temporal decision systems.
mappedRobotics and Embodied AI
Elective branch covering perception planning control and physical interaction.
mappedEdge Model Optimization
How compression runtimes and hardware co-design enable constrained inference.
labTokens and Tokenization
How model inputs become discrete identifiers and why token boundaries affect cost and meaning.
lessonEmbeddings
How learned vectors encode useful similarity and support representation and retrieval.
lessonTransformer Architecture
How attention feed-forward blocks residual paths and positions transform sequences.
lessonTraining versus Inference
How parameter learning differs from runtime generation and serving operations.
labContext Windows
How finite input and output budgets constrain attention state relevance and cost.
labSampling and Decoding
How logits temperature top-p and deterministic choices shape generated sequences.
lessonModel Families and Lifecycle
How capability modality size versioning deprecation and routing affect model selection.
labLatency and Throughput
How queueing prompt processing generation batching and concurrency determine serving performance.
mappedMultimodal Fusion
How models align and combine text image audio and video representations.
lessonMessage Roles and Instruction Priority
How instruction sources and ordered messages establish conversational control context.
lessonPrompt Structure
How clear goals constraints context examples and output contracts guide model behavior.
lessonStructured Outputs
How schemas convert probabilistic text generation into validated application contracts.
lessonTool Calling
How models propose typed actions while applications retain execution authority and validation.
lessonContext Engineering
How systems select assemble order compress and isolate information for each model call.
lessonState and Memory
How applications persist working state history summaries facts and user-controlled records.
lessonExact and Semantic Caching
How reusable computations reduce latency and cost while introducing freshness and correctness risks.
labRAG versus Fine-Tuning
How knowledge access behavior adaptation and tool use solve different system problems.
labIngestion and Chunking
How parsing normalization segmentation metadata and updates create retrievable units.
labVector Search
How embeddings similarity indexes filters and recall tradeoffs retrieve semantic candidates.
labHybrid Search and Reranking
How lexical semantic and learned ranking stages improve candidate precision and recall.
lessonGrounded Generation and Citations
How answer synthesis constrains claims to evidence and preserves inspectable attribution.
mappedKnowledge Provenance
How origin ownership version and transformation metadata make knowledge auditable.
mappedAgent Control Loop
How bounded observe decide act and verify cycles produce controlled autonomy.
mappedWorkflow Orchestration
How explicit state transitions coordinate repeatable AI and software work.
mappedContext and Tool Protocols
How standardized discovery invocation and data exchange connect model systems.
mappedHuman Approval Boundaries
How consequence and reversibility determine when human authorization is mandatory.
mappedAI Feedback and Correction
How interfaces expose progress uncertainty evidence and correction controls.
mappedStreaming AI Interfaces
How incremental transport cancellation state and recovery create responsive experiences.
labLLM Evaluation
How task definitions datasets metrics rubrics judges and experiments measure system quality.
lessonGrounding and Hallucination
How unsupported claims arise and how evidence constraints and verification reduce them.
mappedAI Tracing
How correlated model retrieval tool and application spans expose system behavior.
labCost Latency and Reliability
How budgets fallbacks retries routing and service targets balance operating outcomes.
lessonSecurity and Privacy for LLM Systems
How untrusted inputs sensitive data tools and external knowledge expand the threat model.
mappedSafety Evaluation and Response
How adversarial tests policy checks monitoring and response control harmful behavior.
mappedAI Risk Governance
How ownership classification documentation and review control lifecycle risk.
mappedInference Serving Architecture
How routing batching caching autoscaling and accelerators serve model workloads.
mappedAI Unit Economics
How per-task value quality compute tokens and operational costs determine viability.
mappedCapability-Problem Fit
How uncertain model capabilities map to valuable testable user outcomes.
mappedEvidence Synthesis
How primary sources experiments and explicit confidence create durable knowledge.
mappedInformation Entropy
How entropy and information measures describe uncertainty compression and representation.
researchedPredictive Processing
What hierarchical prediction and prediction error propose about perception and what AI engineers must not infer.
mappedCognitive Architectures
How cognitive models decompose memory reasoning attention and action without becoming literal brain replicas.
mappedLanguage and Thought
How language structures communication and reasoning without equating fluent generation with complete understanding.
mappedPhilosophy of Intelligence
How competing definitions of intelligence change system claims evaluation and responsible communication.
mappedEvidence and Replication
How to distinguish a result from a durable claim through replication synthesis and explicit uncertainty.
mappedCausal Inference
How interventions assumptions and identification separate causal questions from predictive accuracy.
mappedReinforcement Learning
How policies learn sequential decisions from feedback and where reward design and distribution shift fail.
mappedFull-Stack AI Systems
How interfaces APIs orchestration data and model providers form one observable product system.
researchedInference Engine Architecture
How schedulers KV-cache managers model runners and distributed executors determine serving behavior.
mappedMLOps Lifecycle
How versioned artifacts evaluation promotion monitoring rollback and retirement make model changes governable.
mappedDistributed AI Systems
How computation communication memory placement and failure domains shape large-scale AI platforms.
mappedAI System Testing
How to combine deterministic software tests model evaluations adversarial checks and monitored production evidence.
mappedOrganizational AI Change
How capability ownership workflow redesign governance and feedback determine whether AI adoption produces value.
researchedClinical Decision Support
How AI recommendations enter clinical workflows and why evidence oversight usability and escalation determine safety.
mappedAI in Financial Risk
How AI changes fraud risk and underwriting workflows under model-risk and consumer-protection constraints.
mappedAI-Assisted Learning
How AI can support practice feedback and synthesis without replacing learner effort or reliable assessment.
mappedAI for Scientific Discovery
How AI can accelerate search and experiment cycles while preserving measurement validity and reproducibility.
mappedAI-Assisted Software Engineering
How AI changes software throughput review burden system understanding and security risk.
mappedIndustrial AI Automation
How AI interacts with physical constraints safety cases latency and operational recovery in industry.
mappedAI for Climate and Energy
How forecasting optimization and sensing support resource decisions while computation and rebound costs remain visible.
mappedGenerative Media Systems
How models production tools rights provenance and human direction combine in responsible media systems.
mappedAI in Law and Public Services
How automation affects accountable decisions rights access explanation appeal and institutional legitimacy.
mappedAI Labor and Society
How AI changes tasks bargaining power institutions distribution and competitive dynamics beyond headline job counts.
lessonLLM Systems Foundations
Build a connected mental model from tokens through retrieval evaluation security and operations.
labTokenization and Context Window Visualizer
Inspect deterministic token approximations and allocate a finite context budget.
labSampling and Decoding Explorer
Explore how temperature and top-p reshape a fixed token distribution.
labPrompt RAG Fine-Tuning or Tool-Use Decision Lab
Match system symptoms and constraints to an appropriate intervention strategy.
labRAG Pipeline and Retrieval Parameter Visualizer
Tune chunk size retrieval breadth hybrid weight and reranking in a deterministic pipeline.
labLLM Cost Latency and Reliability Simulator
Model request economics latency percentiles and fallback effects under disclosed assumptions.
labEvaluation Strategy Designer
Assemble a balanced evaluation plan from risks cases metrics review and release gates.
guideLLM Application Boundaries
Separate model behavior, application policy, tools, state, and validation into testable system boundaries.
guideRetrieval System Design
Trace ingestion, chunking, search, reranking, generation, citations, and evaluation as one evidence pipeline.
guideAgent Control Paths
Design tool authorization, state, memory, failure handling, and human control around probabilistic decisions.
guideReliable AI Operations
Connect evaluation, latency, throughput, cost, privacy, security, monitoring, and release decisions.
researchedGrounded Research Assistant
Build an evidence-carrying RAG workflow with hybrid retrieval, reranking, citations, evaluation, and abstention.
researchedGoverned Tool-Using Agent
Build an agent with schema-validated tools, independent authorization, state, retries, audit events, and human approval.
researchedAI Support Workspace
Connect a streaming interface, conversation state, retrieval, tool execution, observability, and operating-cost controls.