Project 42
Start learning

Instructor-led Preview

The classroom, on demand.

The same material, taught rather than read. A virtual instructor works through each module on video, with captions and a full transcript, so you can watch a lesson instead of reading one. Same Focus Areas, same courses, same knowledge checks, same sources, and one record either way.

1 lesson filmed so far out of 40 written for the classroom, across 3 of 13 paths. The Focus Areas and their courses are identical to the self-paced curriculum. Every module is already available to read, and anything you finish now carries straight over when its lesson is published.

Focus Area 01

AI Literacy & Foundations

Core mental models, language model generation, prompt anatomy, context tokens, verification, and privacy without assuming technical experience.

01
beginner16 scripted · ~288 min

AI Foundations

Build a clear mental model of AI, prompts, and agents without assuming technical experience.

For people starting with ai or filling foundational gaps.
1. What AI Does—and Does Not DoScripted · filming soon
2. AI Systems and Useful WorkScripted · filming soon
3. How Language Models Produce ResponsesScripted · filming soon
4. Context, Tokens, and ModalitiesScripted · filming soon
5. Privacy, Safety, and Responsible UseScripted · filming soon
6. Prompt with PurposeScripted · filming soon
7. Prompt Anatomy and Success CriteriaScripted · filming soon
8. Context and Evidence ConstructionScripted · filming soon
9. Examples and Output ContractsScripted · filming soon
10. Verification and Iterative ImprovementScripted · filming soon
11. Research with EvidenceScripted · filming soon
12. Writing and Transformation WorkflowScripted · filming soon
13. Coding and Analysis WorkflowScripted · filming soon
14. Safe Tool-Use WorkflowScripted · filming soon
15. Agents, Tools, and GuardrailsWatch lesson →
16. AI Foundations CapstoneScripted · filming soon
Watch published lesson
07
beginnerNot scripted yet

What Is Agentic AI?

Classify models, assistants, workflows, agents, agentic systems, and multi-agent systems from observable evidence instead of product labels.

For learners from beginner through advanced who need a precise, provider-neutral way to explain and evaluate agentic ai.
1. Name the Layers in an AI ExperienceRead module →
2. Classify Who Chooses the Next StepRead module →
3. Locate Tools, State, and Human AuthorityRead module →
4. Compare Exact Product ExperiencesRead module →
5. Defend an Agentic-AI ClassificationRead module →
Read path online

Focus Area 02

Developer & Practitioner AI

Provider-neutral evaluation, capability comparison, structured outputs, function calling, and hands-on practice across Anthropic, OpenAI, and Google Gemini.

02
intermediateNot scripted yet

Providers in Practice

Learn what transfers across providers and where Anthropic, OpenAI, and Google workflows differ.

For practitioners choosing tools or supporting a mixed-provider team.
1. Anthropic in PracticeRead module →
2. OpenAI in PracticeRead module →
3. Choose a Provider DeliberatelyRead module →
4. Compare AI Providers without False EquivalenceRead module →
5. Plan a Cross-Provider MigrationRead module →
6. Execute, Roll Back, and Close a Provider MigrationRead module →
Read path online
03
intermediateNot scripted yet

Claude in Practice

Build practical Claude workflows from surface selection and prompting through APIs, tools, safety, evaluation, and migration.

For learners and practitioners who want to use claude deliberately in individual, developer, and agent workflows.
1. Navigate the Claude EcosystemRead module →
2. Prompt Claude with Testable ContractsRead module →
3. Build Reliable Claude API WorkflowsRead module →
4. Build Bounded Claude Tool and Agent LoopsRead module →
5. Secure Claude Workflows with Layered Trust BoundariesRead module →
6. Evaluate and Observe Claude WorkflowsRead module →
7. Migrate Claude Workflows with EvidenceRead module →
Read path online
04
intermediateNot scripted yet

OpenAI in Practice

Choose the right OpenAI surface and build evaluated, secure, observable, migration-ready API, tool, and Codex workflows.

For learners and practitioners using chatgpt, the openai api, or codex for individual, application, and development workflows.
1. Navigate the OpenAI EcosystemRead module →
2. Prompt OpenAI Models with Evaluated ContractsRead module →
3. Build Reliable OpenAI Responses API WorkflowsRead module →
4. Build Bounded OpenAI Tool and Codex Agent LoopsRead module →
5. Secure OpenAI Workflows with Layered Trust BoundariesRead module →
6. Evaluate and Observe OpenAI WorkflowsRead module →
7. Migrate OpenAI Workflows with EvidenceRead module →
Read path online
05
intermediateNot scripted yet

Gemini in Practice

Choose the right Gemini surface and build evaluated, secure, observable, migration-ready API, tool, and agent workflows.

For learners and practitioners using gemini apps, google ai studio, or the gemini api for individual, prototype, and application workflows.
1. Navigate the Gemini Ecosystem and InterfacesRead module →
2. Prompt Gemini with Evaluated ContractsRead module →
3. Build Reliable Gemini API WorkflowsRead module →
4. Build Bounded Gemini Tool and Agent LoopsRead module →
5. Secure Gemini Workflows with Layered Trust BoundariesRead module →
6. Evaluate and Observe Gemini WorkflowsRead module →
7. Migrate Gemini Workflows with EvidenceRead module →
Read path online
12
intermediateNot scripted yet

DeepSeek in Practice

Build evaluated, cost-optimized applications using DeepSeek V3, DeepSeek R1 reasoning models, and Mixture-of-Experts architecture.

For developers, system architects, and ai practitioners leveraging open and hosted deepseek models.
1. Navigate the DeepSeek Ecosystem & ArchitectureRead module →
2. DeepSeek R1 Reasoning Tokens & Chain-of-ThoughtRead module →
3. DeepSeek API Workflows, Tool Calling & Cost OptimizationRead module →
Read path online
13
intermediateNot scripted yet

Mistral in Practice

Build applications using Mistral Small/Large, Codestral for code generation, Pixtral for vision, and native embeddings.

For developers and engineers building multilingual, code, and multimodal applications on mistral.
1. Navigate the Mistral Ecosystem & Model FamilyRead module →
2. Mistral Function Calling & Codestral FIMRead module →
3. Pixtral Vision & Mistral EmbeddingsRead module →
Read path online

Focus Area 03

Frontier Agentic Systems & MCP

Bounded agent loops, tool authority, memory boundaries, Model Context Protocol (MCP) architecture, multi-agent handoffs, and scored capstone.

06
intermediate12 scripted · ~192 min

Reliable Agent Workflows

Turn agent capability into dependable work through clear scope, permission boundaries, and evidence-based review.

For people using coding, research, or operations agents for work that must be checked and repeatable.
1. Design a Bounded Agent LoopScripted · filming soon
2. Design Typed Tool Contracts and Action ControlsScripted · filming soon
3. Engineer Context for Reliable DecisionsScripted · filming soon
4. Design Safe Agent Memory BoundariesScripted · filming soon
5. Understand MCP Architecture and ContractsScripted · filming soon
6. Secure MCP Trust BoundariesScripted · filming soon
7. Choose Reliable Orchestration PatternsScripted · filming soon
8. Build Auditable Multi-Agent HandoffsScripted · filming soon
9. Evaluate Agent Behavior with EvidenceScripted · filming soon
10. Observe Agent Quality, Cost, and RiskScripted · filming soon
11. Operate, Recover, and Improve Agent SystemsScripted · filming soon
12. Reliable Agent Capstone: Design, Test, and OperateScripted · filming soon
Read path online

Focus Area 04

Retrieval, RAG & Fine-Tuning

Advanced retrieval architectures, hybrid search, embedding stores, knowledge graphs, and LoRA/QLoRA fine-tuning.

09
advancedNot scripted yet

Advanced RAG Engineering

Design, index, retrieve, rerank, and evaluate production Retrieval-Augmented Generation systems across hybrid search and knowledge graphs.

For ml engineers, backend developers, and data architects building enterprise knowledge systems.
1. Chunking Strategies & Document Ingestion in PracticeRead module →
2. Hybrid Search & Cross-Encoder RerankingRead module →
3. GraphRAG & Multi-Hop Entity ReasoningRead module →
4. RAG Evaluation & The RAG TriadRead module →
5. Advanced RAG Capstone: Design, Implement & EvaluateRead module →
Read path online
10
advancedNot scripted yet

Model Customization & Fine-Tuning

Master Parameter-Efficient Fine-Tuning (LoRA, QLoRA), instruction dataset curation, preference alignment (DPO), and model evaluation.

For machine learning engineers, ai practitioners, and developers tailoring open-weight models.
1. The AI Customization Decision MatrixRead module →
2. LoRA & QLoRA Parameter-Efficient Fine-TuningRead module →
3. Dataset Curation & Direct Preference Optimization (DPO)Read module →
4. Evaluation, Merging & GGUF/AWQ QuantizationRead module →
5. Model Customization Capstone: Train, Align & ServeRead module →
Read path online

Focus Area 05

Self-Hosted, Open-Weight & AIOps

Open-weight model selection, vLLM/Ollama serving, VRAM calculations, artifact integrity, endpoint security, and disaster recovery.

08
advanced12 scripted · ~200 min

Self-Hosted Model Operations

Select, verify, deploy, evaluate, observe, update, roll back, and recover a self-hosted model endpoint.

For practitioners and operators responsible for open-weight or otherwise self-managed model services across workstations, edge, on-premises, and cloud environments.
1. Choose a Deployment Shape and Operating ModelScripted · filming soon
2. Model Identity, Licensing, and ProvenanceScripted · filming soon
3. Model Artifact Integrity and Safe PromotionScripted · filming soon
4. Hardware, Runtime, and Capacity PlanningScripted · filming soon
5. Model Serving and API Compatibility ContractsScripted · filming soon
6. Endpoint Identity, Network, and Secrets SecurityScripted · filming soon
7. Evaluate the Exact Serving BuildScripted · filming soon
8. Observability, Cost, and PerformanceScripted · filming soon
9. Scaling, Failure, and Capacity ControlsScripted · filming soon
10. Model Update and Rollback LifecycleScripted · filming soon
11. Model Incident Response and RecoveryScripted · filming soon
12. Self-Hosted Model Operations CapstoneScripted · filming soon
Read path online

Focus Area 06

AI Security, Red-Teaming & Governance

OWASP Top 10 for LLMs, sandboxing, guardrails, compliance frameworks, and cryptographic audit receipts.

11
advancedNot scripted yet

Adversarial AI Security & Red-Teaming

Identify, exploit, and remediate OWASP Top 10 for LLMs vulnerabilities, prompt injections, and jailbreak vectors.

For security engineers, red-teamers, cloud architects, and ai application developers.
1. OWASP Top 10 for LLM ApplicationsRead module →
2. Jailbreaking, Prompt Injection & Defense in DepthRead module →
3. Automated Red-Teaming & Vulnerability ScannersRead module →
4. Adversarial Security Capstone: Audit, Attack & FortifyRead module →
Read path online

Prefer reading the text-first modules? Explore the self-paced library →