AI Development for Engineers

A practical course for developers who want to add AI to their stack. No deep mathematics — the focus is on real-world application: integrating LLM APIs, building RAG systems, developing AI agents, and shipping AI-powered features in web applications.

Requirements

• 1+ year of JavaScript or Python development experience
• Understanding of REST APIs and basic web development principles
• Successful completion of the test task
• 1–2 hours per day

Programme

LLM Fundamentals

• How GPT, Claude, and Gemini work — what every developer needs to know
• Tokens, context window, temperature — parameters and their effects
• Choosing the right model for the task

Prompt Engineering

• System prompts and roles
• Techniques: few-shot, chain-of-thought, self-consistency
• Structured output: JSON mode, function calling

Working with APIs

• OpenAI SDK and Anthropic SDK: text, vision, embeddings
• Streaming responses
• Tool use / function calling in practice

RAG — AI on Your Own Data

• Embeddings and semantic search
• Vector DBs: Pinecone, Chroma, pgvector
• Building a pipeline: load → chunk → retrieve → generate

AI Agents

• Agent architecture: planning, tool use, memory
• LangChain and LlamaIndex — overview and practical use
• Multi-step reasoning in practice

Integration into Web Applications

• Streaming UI: Server-Sent Events, WebSocket
• UX patterns for AI features
• Examples: chatbot, document search, autocomplete

Evaluation and Safety

• Hallucinations: causes and mitigation strategies
• Prompt injection and defences
• Cost management, rate limits, request caching

Format

The course is individual and conducted online with weekly sessions. Includes practical assignments after each section. Final project: an AI feature built into a real application. A certificate of completion is issued upon passing.

Instructors

Engineers with hands-on experience shipping AI features in production projects at Smartym Pro.