-
Phase 1. Who the product engineer is · 3 questions
The role and the work cycle: from user problem to outcome metric, how it differs from a developer and a PM, the skills map.
Take the phase quiz → self-check after reading the phase
-
Phase 2. How AI works · 7 questions
The foundation: how models predict text, hallucinations, tokens and cost, context, tool calling and agents.
- How AI Models Work: Predicting the Next Token
- AI Hallucinations: Why the Model Confidently Makes Things Up
- Tokens and Cost: What You Pay For When Working With AI
- Context: What the Model Holds in Mind at Once
- Tool Calling: How the Model Reaches Beyond Text
- Agents: A Model in a Loop with Tools and a Goal
Take the phase quiz → self-check after reading the phase
-
Phase 3. Building LLM applications · 6 questions
How to embed AI into your own product: what an LLM feature is made of, orchestration with LangChain, RAG and embeddings, vector databases, and agent applications.
- What an LLM feature is made of
- LangChain: orchestrating LLM applications
- RAG and embeddings: giving the model your data
- Vector databases
- Agent applications: tools and autonomous loops
Take the phase quiz → self-check after reading the phase
-
Phase 4. Working with agents · 8 questions
Practice: the session loop, understanding a codebase, development via plans, debugging, review and testing, configuring the agent — skills, MCP, memory.
- Working with Agents: The Basic Loop of a Productive Session
- Understanding a Codebase with an Agent
- Feature Development: Brainstorming and Plans with an Agent
- Finding and Fixing Bugs with an Agent
- Reviewing and Testing Code with an Agent
- Configuring Agents: How to Make the Agent Know Your Project
- Skills: packing a methodology into an AI agent's skill
- MCP: How to Give an AI Agent Access to Tools and Data
- Memory Bank: Persistent Memory of an AI Agent for a Project
Take the phase quiz → self-check after reading the phase
-
Phase 5. Start with the problem · 6 questions
Problem before solution, talking to users, outcome metrics, the smallest valuable slice.
- The Problem, Not the Solution
- Contact with the User
- Outcome Metrics, Not Output Metrics
- Prioritization: the smallest valuable slice
Take the phase quiz → self-check after reading the phase
-
Phase 6. Domain model and ontologies · 6 questions
A system of concepts before code: ontology in plain words, conceptual model vs DB schema, the ubiquitous language and context boundaries, the path from model to an agent-ready spec, and knowledge graphs for LLMs.
- A system of concepts: ontology in plain words
- A conceptual model is not a database schema
- The ubiquitous language: a domain vocabulary as a contract
- From a concept model to a spec for AI
- Ontologies and knowledge graphs for LLMs
Take the phase quiz → self-check after reading the phase
-
Phase 7. Contract and tooling for AI · 5 questions
How to give the agent a task: a language for AI and slice → contract.
- Which Programming Language to Choose for AI Coding
- From a Product Slice to a Contract an AI Can Build
Take the phase quiz → self-check after reading the phase
-
Phase 8. Build and verify · 6 questions
Reviewing AI code, the executable standard vs linters, accepting the result against criteria and tests from the spec.
- How to Review Code Written by AI
- Executable agent rules vs SonarQube vs ESLint vs tech lead review
- Accepting AI output: acceptance criteria and tests from the spec
Take the phase quiz → self-check after reading the phase
-
Phase 9. Ownership and outcome · 6 questions
Ownership from idea to user, AI as leverage, shipping and the metrics loop as one person, and which methodology ties the whole path together.
- Ownership from idea to user
- AI as the product engineer's leverage
- Shipping and the Metrics Loop, Solo
- A Methodology for AI: Use Case Pattern and Other Ways
Take the phase quiz → self-check after reading the phase
-
Phase 10. Development methodologies and process · 8 questions
Waterfall, Agile, Scrum, Kanban, XP, estimation and scaling — how a team organizes work from idea to release.
- Development Models: From Waterfall to Agile
- Scrum: Roles, Events, and Artifacts
- Kanban: Flow, WIP Limits, and Pull
- Extreme Programming: the Engineering Practices of Agile
- Estimation and planning: story points and velocity
- Scaling Agile: SAFe and LeSS
Take the phase quiz → self-check after reading the phase
Программа обучения v1.3.0
Product Engineer
How one person with AI ships a whole product — from the problem to outcome metrics. Not about the craft (AI knows that from backend/frontend), but the product-engineering layer: product thinking, a contract for AI, verifying the result, and owning the outcome.
The program has 61 self-check questions in total.