Frontier Foundations. How frontier labs hire, the fundamentals every round assumes, the progressive coding screen, reading the labs, and the values round. For: everyone, before any track.
Foundations map
Learn the loop.Then pass it.
- 0%
- complete
- 0/71
- graded tasks done
- 72h
- to go
Start here
C1 · The hiring loop
The Loop, Round by Round
Modules
0 of 163 lessons- 0/6C1·The hiring loopIn progress
How Frontier Labs Hire
The loop at Anthropic, OpenAI and Perplexity, round by round: what each round tests, what the AI-use rules are, and why the written application is your first interview.
Nadia Okafor, Interview coach6 lessons · 1.7h - 0/8C2·FundamentalsFull course
LLM Fundamentals Interviewers Probe
Tokens, context windows, sampling, tool use, caching, structured outputs, model choice and cost math: the fundamentals every applied round assumes.
Nadia Okafor, Interview coach8 lessons · 3.4h- ReadTokens, Context Windows and What They Cost20mLocked: part of the full course
- TaskBuild a Cost Calculator30mLocked: part of the full course
- ReadThinking, Effort and Why Outputs Vary20mLocked: part of the full course
- ReadTool Use and Structured Outputs26mLocked: part of the full course
- TaskPick the Model30mLocked: part of the full course
- ReadPrompt Caching and Latency22mLocked: part of the full course
- WriteWrite It: A Cost and Caching Plan40mLocked: part of the full course
- QuizModule C2 Quiz14mLocked: part of the full course
- 0/6C3·Coding screenFull course
Progressive Coding Drills
The multi-level coding screen: design the data model first, extend without rewriting, and stay calm against the clock.
Nadia Okafor, Interview coach6 lessons · 3.5h- ReadHow Progressive Coding Screens Work15mLocked: part of the full course
- DrillDrill: A Feature-Flag ServiceFree preview90m
- ReadData Models That Survive Level 416mLocked: part of the full course
- TaskRanking and Formatting Warm-Up20mLocked: part of the full course
- DrillDrill: An LLM Gateway Ledger60mLocked: part of the full course
- QuizModule Quiz: Progressive Coding10mLocked: part of the full course
- 0/7C4·Reading the labsFull course
Reading the Labs
How to read and critique what lab leaders publish: steelman, assumptions, strongest counterargument, and what would change your mind.
Nadia Okafor, Interview coach7 lessons · 2.7h- ReadHow to Read a Lab Essay16mLocked: part of the full course
- ReadEssay: Machines of Loving Grace25mLocked: part of the full course
- ReadEssay: The Adolescence of Technology25mLocked: part of the full course
- ReadThe Policies Behind the Essays20mLocked: part of the full course
- WriteWrite It: Critique an Essay40mLocked: part of the full course
- ReadStudy Guide: Engineering Posts and Other Labs' Documents22mLocked: part of the full course
- QuizC4 Quiz: Reading the Labs12mLocked: part of the full course
- 0/10C5·Values roundFull course
Values & Behavioral
The round that reportedly fails the most candidates: mission versus money, disagreement, failure and safety tradeoffs, practised live.
Nadia Okafor, Interview coach10 lessons · 3.5h- ReadWhat the Values Round Tests15mLocked: part of the full course
- ReadBuilding Your Story Bank15mLocked: part of the full course
- WriteWrite It: Two Stories30mLocked: part of the full course
- ReadAnswering Mission Questions Honestly15mLocked: part of the full course
- TalkLive Practice: A Values InterviewFree preview30m
- TalkLive Practice: The Project Deep Dive30mLocked: part of the full course
- ReadOpeners: Your Pitch, Why You're Moving, and Your Questions16mLocked: part of the full course
- ReadMapping Your Stories to Published Values12mLocked: part of the full course
- TalkLive Practice: The LLM Project Deep Dive35mLocked: part of the full course
- QuizModule C5 Quiz12mLocked: part of the full course
- 0/18C6·Fundamentals roundFull course
How LLMs Work: Model Literacy
Transformers, attention, the KV cache, prefill versus decode, tokenization, sampling, the training pipeline, hallucination, prompt injection, evals and embeddings: the why behind model behaviour that fundamentals and design rounds probe, vendor-neutral where interviews are.
Mira Vasquez, Fundamentals-round interviewer (fictional practice persona)18 lessons · 6.8h- ReadTransformers for Applied Engineers16mLocked: part of the full course
- TaskAttention by Hand and a Latency Estimator30mLocked: part of the full course
- ReadTokenization: How Text Becomes Tokens15mLocked: part of the full course
- TaskBuild a Toy BPE Tokenizer40mLocked: part of the full course
- ReadSampling: Logits, Temperature and Why Zero Isn't Deterministic14mLocked: part of the full course
- TaskTemperature, Top-p and a Seeded Sampler30mLocked: part of the full course
- ReadTraining: How a Base Model Becomes an Assistant18mLocked: part of the full course
- ReadHallucination: Causes, Detection and Mitigation16mLocked: part of the full course
- TaskA Citation-Grounding Checker35mLocked: part of the full course
- ReadPrompt Injection and Agent Safety Fundamentals16mLocked: part of the full course
- TaskHarden a Tool Loop Against an Injected Document40mLocked: part of the full course
- ReadEvals 101: Measuring a Model You Can't Unit-Test16mLocked: part of the full course
- TaskA Small Eval Harness with an LLM Judge35mLocked: part of the full course
- ReadEmbeddings and Retrieval 10115mLocked: part of the full course
- TaskBrute-Force Cosine Top-k and a Chunker20mLocked: part of the full course
- ReadVendor-Neutral Vocabulary: Speak Every Lab's Language12mLocked: part of the full course
- TalkRole-play: The Fundamentals Round25mLocked: part of the full course
- QuizModule C6 Quiz12mLocked: part of the full course
- 0/16C7·Cindervale Freight (fictional)Full course
The Modern Agent Stack: MCP, Sub-Agents, Skills and Claude Code
Build and secure an MCP server, fan work out to cheap sub-agents and prove it paid off, package a playbook as an Agent Skill, keep long agents append-only under preserved thinking, red-team an agent, and defend a build-or-buy platform decision to a CTO.
Rhea Petrakis, Head of Platform Engineering16 lessons · 8.1h- ReadThe Agent Stack Map: Who Supplies the Harness, Who Supplies the Deployment16mLocked: part of the full course
- ReadMCP Fundamentals: Tools, Resources, Prompts and the Wire18mLocked: part of the full course
- TaskBuild an MCP Server for Shipments50mLocked: part of the full course
- TaskSecure the MCP Server45mLocked: part of the full course
- WriteWrite It: The CISO Security Brief for an MCP Server40mLocked: part of the full course
- ReadSub-Agents in Practice: When Splitting Pays and When It Doesn't14mLocked: part of the full course
- TaskFan Out a Contract Review, Then Prove It Paid Off50mLocked: part of the full course
- ReadAgent Skills: Packaging Know-How for Progressive Disclosure14mLocked: part of the full course
- TaskPackage the Exception Playbook as a Skill45mLocked: part of the full course
- ReadContext Engineering for Long-Running Agents18mLocked: part of the full course
- TaskCut a 40-Turn Trace Without Touching History45mLocked: part of the full course
- ReadClaude Code for FDEs: Your Tool, and Your Customer's Rollout15mLocked: part of the full course
- ReadComputer Use: When GUI Automation Beats an API12mLocked: part of the full course
- TaskRed-Team an Agent, Then Defend It45mLocked: part of the full course
- TalkDesign Interview: Build or Buy the Agent Platform45mLocked: part of the full course
- QuizModule Quiz: The Modern Agent Stack15mLocked: part of the full course
- 0/11C8·Practice review panel (fictional)Full course
Portfolio Capstone
The Foundations portfolio capstone: build a real project against the Claude API on your own key, in a domain you choose, starting from the capstone-kit repository template. Run four real-model labs, write the README, decision record and eval report, then defend it to a practice reviewer who has read them.
Ondine Corriveau, Staff engineer and portfolio reviewer11 lessons · 4.6h- ReadWhy a Portfolio Project12mLocked: part of the full course
- ReadScoping Your Capstone in a Weekend15mLocked: part of the full course
- ReadReal-Model Labs: How to Run Them on Your Key14mLocked: part of the full course
- WriteLab 1 Report: Prompt Iteration on a Real Model25mLocked: part of the full course
- WriteLab 2 Report: Retrieval Eval20mLocked: part of the full course
- WriteLab 3 Report: Calibrating an LLM Judge25mLocked: part of the full course
- WriteLab 4 Report: Classifier Cost, Latency and Refusals20mLocked: part of the full course
- WriteCapstone README and Decision Record60mLocked: part of the full course
- WriteThe Eval Report40mLocked: part of the full course
- TalkDefend It: Capstone Review with Ondine35mLocked: part of the full course
- QuizModule C8 Quiz12mLocked: part of the full course
- 0/15C9·Ossian Careers (fictional)Full course
Getting the Job
Career mechanics around the loop: choose the role your evidence supports, decode postings, write resume bullets and outreach notes that survive the screen, handle the recruiter screen, take-homes and the online assessment, and understand levels, equity and negotiation. Every claim is labelled; general hiring practice is marked General.
Hana Ostrowski, Career Coach, Applied AI Roles15 lessons · 5.1h- ReadChoosing Your Role14mLocked: part of the full course
- TaskDecode the Posting30mLocked: part of the full course
- ReadResume Bullets That Survive the Screen15mLocked: part of the full course
- WriteWrite It: Rewrite Three Resume Bullets30mLocked: part of the full course
- ReadLinkedIn and Public Proof12mLocked: part of the full course
- WriteReferrals and Outreach (Write the Note)30mLocked: part of the full course
- ReadThe Application and "Why Us?" Answers12mLocked: part of the full course
- TalkRole-play: The Recruiter Screen30mLocked: part of the full course
- WriteTake-Homes and Case Studies (Review a README)40mLocked: part of the full course
- ReadThe Online Assessment Environment10mLocked: part of the full course
- ReadLevels, Compensation and Equity Basics16mLocked: part of the full course
- TalkRole-play: Negotiating the Offer30mLocked: part of the full course
- ReadTeam Matching, Offers, Reapplying and Timing12mLocked: part of the full course
- ReadOther Employers: Google DeepMind and FDE-Heavy Companies12mLocked: part of the full course
- QuizGetting the Job: Module Quiz10mLocked: part of the full course
- 0/16C10·Thistledown Outdoor Co. (fictional)Full course
Agentic System Design
The design-level synthesis of the agent lessons: a method for any 'design an agent' question, workflow patterns versus agents, the agent loop's tools, exits, budgets and loop detection, multi-agent architectures and their token bill, state, memory and context budgets, idempotency, checkpoints and human approval, least privilege and approval tiers, and evaluating and operating agents, for the system design round and for a real customer deployment.
Lucia Ferreira, Director of Automation16 lessons · 7.3h- ReadA Design Method for Agent Questions20mLocked: part of the full course
- ReadWorkflow Patterns: Chaining, Routing, Parallelization, Orchestrator-Workers and Evaluator-Optimizer20mLocked: part of the full course
- TaskA Pattern Router for Automation Requests30mLocked: part of the full course
- ReadThe Agent Loop in Depth: Tools, Stop Conditions, Budgets and Loop Detection20mLocked: part of the full course
- TaskA Bounded Agent Loop with Loop Detection35mLocked: part of the full course
- ReadMulti-Agent Architectures: Orchestrators, Handoffs and the Token Bill20mLocked: part of the full course
- TaskSingle Agent or Orchestrator? A Cost Estimator30mLocked: part of the full course
- ReadState, Memory and Context Engineering20mLocked: part of the full course
- ReadReliability and Recovery: Idempotency, Checkpoints, Retries and Human Approval20mLocked: part of the full course
- ReadSafety, Permissions and Blast Radius20mLocked: part of the full course
- TaskAn Approval-Gate Policy Engine with an Audit Log35mLocked: part of the full course
- ReadEvaluating and Operating Agents20mLocked: part of the full course
- TalkDesign Interview: A Support Agent That Can Issue Refunds45mLocked: part of the full course
- TalkDesign Interview: A Research Agent for Analysts45mLocked: part of the full course
- WriteWrite It: Agent or Workflow? A Design Doc for Warranty Claims40mLocked: part of the full course
- QuizModule Quiz: Agentic System Design15mLocked: part of the full course
- 0/26C11·Larkspur Bank (fictional)Full course
Agent Harness Engineering
Build a coding-agent harness from scratch, the way Claude Code and the Claude Agent SDK are built: the full agent loop with parallel tool calls and every stop reason, a tool registry and runtime, a file edit tool, layered context with compaction and memory, permission modes and hooks, planning and subagents, budgets, checkpoints and resume, tracing and cost, harness evals, defence against injection through tool results, and how to deploy and operate a harness for a customer.
Mira Halvorsen, Director of Developer Platform26 lessons · 13.3h- ReadWhat a Harness Is: The Seven Layers Around the Model22mLocked: part of the full course
- ReadThe Loop in Full: Every Stop Reason, Parallel Calls and the History Rules24mLocked: part of the full course
- TaskWren's Core Loop40mLocked: part of the full course
- ReadTool Design and Execution: The Runtime Between the Model and the World22mLocked: part of the full course
- TaskWren's Tool Registry35mLocked: part of the full course
- ReadThe Built-in Toolkit: File Editing, Bash, Search and Web Fetch22mLocked: part of the full course
- TaskWren's File Edit Tool40mLocked: part of the full course
- ReadContext Assembly: Layered Prompts, Memory Files, Skills and a Cache-Friendly Layout22mLocked: part of the full course
- ReadLong Runs: Compaction, Context Editing and Memory22mLocked: part of the full course
- TaskCompaction That Never Drops the Task40mLocked: part of the full course
- ReadPermissions and Hooks: Enforcing Policy in Code22mLocked: part of the full course
- TaskA Permission Engine and Hooks in the Loop40mLocked: part of the full course
- ReadPlanning, Subagents and Budgets22mLocked: part of the full course
- TaskA todo_write Planner and an Isolated Subagent45mLocked: part of the full course
- ReadDurability: Checkpoints, Idempotency, Interrupt and Resume22mLocked: part of the full course
- TaskCheckpoint, Crash and Resume Without Repeating Writes40mLocked: part of the full course
- ReadObservability and Replay: Traces, Cost and Debugging a Recorded Run25mLocked: part of the full course
- TaskWren's Tracer, with Tokens, Dollars and Redaction40mLocked: part of the full course
- ReadEvaluating a Harness: Task Suites, pass^k, Cost per Solved Task and Regression Gates25mLocked: part of the full course
- ReadSecurity: Tool Results Are Untrusted25mLocked: part of the full course
- TaskInjection Defence, with Untrusted Wrapping, a Trifecta Block and an Output Scanner40mLocked: part of the full course
- ReadDeploying a Harness for a Customer: Build or Buy, Tenancy, Rollout, SLOs and Cost25mLocked: part of the full course
- TalkDesign Interview: A Coding-Agent Harness for Larkspur Bank45mLocked: part of the full course
- TalkWar Room: The Harness Is Misbehaving30mLocked: part of the full course
- WriteWrite It: A Harness Architecture Doc for Larkspur's Architecture Board45mLocked: part of the full course
- QuizModule Quiz: Agent Harness Engineering20mLocked: part of the full course
- 0/24C12·Halden Manufacturing (fictional)Full course
Enterprise Agents End to End
Build one enterprise agent pilot end to end: chatbot vs workflow vs agent, comparing models on a business task, no-code agents in n8n, RAG with citations and the LLM-wiki pattern, scoped tools over APIs and MCP with real authorization, state-graph orchestration (LangGraph) with typed routing and confidence-based escalation, a headless ERP with version-bound human approvals, evaluation and tracing with Langfuse, durable execution with Temporal, and deploying, operating and presenting the pilot. Ends with a complete working solution you can download and run.
Tobias Lindqvist, Head of Procurement Operations24 lessons · 12.3h- ReadChatbot, Workflow or Agent? Choosing the Shape of an Enterprise Pilot22mLocked: part of the full course
- ReadComparing GPT, Claude and Gemini on the Same Business Task22mLocked: part of the full course
- TaskA Model Scorecard for Supplier Onboarding35mLocked: part of the full course
- ReadNo-Code Agents with n8n, and When to Move to Code24mLocked: part of the full course
- ReadEnterprise RAG with Citations, Access Boundaries and Live Data24mLocked: part of the full course
- ReadThe LLM Wiki Pattern: Index-Based Retrieval Without Embeddings22mLocked: part of the full course
- TaskWiki vs RAG Retrieval on Halden's Policies40mLocked: part of the full course
- ReadTools, APIs and MCP Authorization: The Model Suggests, the Service Decides26mLocked: part of the full course
- TaskScoped Supplier Tools with Server-Side Authorization45mLocked: part of the full course
- ReadOrchestration with State Graphs24mLocked: part of the full course
- ReadTyped Routing and Decision Models24mLocked: part of the full course
- TaskA Resumable Graph and a Routing Eval45mLocked: part of the full course
- ReadHeadless ERP and Human Approvals25mLocked: part of the full course
- TaskAn Approval-Gated Supplier Workflow45mLocked: part of the full course
- ReadEvaluation and Observability with Langfuse25mLocked: part of the full course
- TaskAn Eval Scorecard and a Regression Gate40mLocked: part of the full course
- ReadDurable Execution with Temporal25mLocked: part of the full course
- TaskReceipt-Checked Retries and a Replay Check40mLocked: part of the full course
- ReadDeploying and Operating the Pilot24mLocked: part of the full course
- ReadThe End-to-End Solution Walkthrough25mLocked: part of the full course
- TalkDesign Interview: Halden's Procurement Agent Pilot45mLocked: part of the full course
- TalkPilot Review: Presenting the Results to the Business Owner30mLocked: part of the full course
- WriteWrite It: A Rollout Proposal for the Procurement Agent40mLocked: part of the full course
- QuizModule Quiz: Enterprise Agents End to End20mLocked: part of the full course
Unlock every track.The first lessons of every track are free. Get all four tracks, every graded task and three mock loops with a one-time payment.
Get full access · $249