AI Engineer
Slava Skryabin
VP of Engineering at one of the largest international IT companies, Slava is a seasoned automation architect with a US patent in automation architecture solutions. He has designed and built advanced test automation frameworks from scratch using Java, Python, Ruby, Scala, and JavaScript — covering UI, mobile, web services, databases, and even security vulnerability scanning.
In recent years, Slava has also been actively involved in AI engineering, applying machine learning and AI-powered solutions to enhance in development, test automation, anomaly detection, and intelligent data analysis.
A proud graduate of Portnov Computer School, Slava returned as an instructor and has been teaching various courses at the school for over 15 years. He is the creator of several popular programs and is especially well-known among students for the Silver Bootcamp with Java, one of the school’s most in-demand courses.
In recent years, Slava has also been actively involved in AI engineering, applying machine learning and AI-powered solutions to enhance in development, test automation, anomaly detection, and intelligent data analysis.
A proud graduate of Portnov Computer School, Slava returned as an instructor and has been teaching various courses at the school for over 15 years. He is the creator of several popular programs and is especially well-known among students for the Silver Bootcamp with Java, one of the school’s most in-demand courses.
What You’ll Master in This Bootcamp
The AI Engineer curriculum is structured into four focused modules.
- AI Agentic Development
- Claude Code — Introduction, installation, and setup for agentic coding in the terminal and IDE, plus an ideal agentic project directory structure.
- Spec-Driven Development — Use structured specs to get production-ready code on the first pass (Spec Kit, Superpowers, etc.).
- Model Context Protocol (MCP) — Connect to Jira, Playwright, and Chrome DevTools MCPs to go from ticket → code → tested feature end-to-end.
- Agent Skills — Build AI Skills and package reusable capabilities as SKILL.md files to automatically apply the right expertise to the right task.
- Agentic Teams — Delegate long-horizon tasks to specialized sub-agents (Architect, Developer, Code Reviewer, QA Tester) running in isolated context windows and orchestrated under a main agent.
- Agentic Coding — Ship features with CLAUDE.md, DESIGN.md, AGENTS.md, hooks, slash commands, and rules.
- Hands-On Bootcamp Projects — Build multiple end-to-end agentic applications with AI-driven quality testing from scratch, applying Claude Code, MCP, Skills, and sub-agents to real-world use cases you can add to your portfolio.
- Prompt Engineer with Python
- Python for AI & Data Handling — Learn Python from the ground up, including data processing, APIs, error handling, and AI model QA automation.
- Prompt Engineering & In-Context Learning — Master zero-shot (ZSL), few-shot (FSL), chain-of-thought (CoT), self-consistency, and tree-of-thought (ToT) prompting.
- Structured Outputs & Data Extraction — Extract structured data from unstructured text (PDFs, documents) and validate it for production pipelines.
- Prompt Optimization & Testing — Evaluate prompts systematically with golden datasets and automated grading—moving from guesswork to measurable quality.
- Generative AI Engineer
- Chaining, LangChain & Retrieval-Augmented Generation (RAG) — Build intelligent chatbots, memory-based assistants, and AI-powered search.
- AI Agents & Tool Integration (ReAct) — Combine reasoning with action to plan and execute tasks via calculators, APIs, and custom tools.
- AI Guardrails: Toxicity & Factuality — Implement safety layers with toxicity filtering, factuality scoring, and robust fallback mechanisms.
- AI QA Testing & Monitoring — Add test coverage for data pipelines, model performance, and LLM outputs to ensure production reliability.
- Model Context Protocol (MCP) — Build MCP servers and clients in Python to connect Claude to apps, databases, and internal APIs.
- Agent Evaluation & Observability — Design evals, tracing, and cost tracking to measure agent reliability in production.
- AI Fluency & Responsible AI — Apply 4D framework (Delegation, Description, Discernment, Diligence) alongside professional and ethical best practices.
- Cloud Deployment (AWS SageMaker & Bedrock) — Take models live with scalable, enterprise-grade inference.
- AI Researcher
- Deep Learning & Model Training (PyTorch) — Train, specialize, and optimize foundation models using Hugging Face Transformers.
- Parameter-Efficient Fine-Tuning (PEFT) with LoRA — Fine-tune large foundation models using Low-Rank Adaptation (LoRA) and adapter methods.
Address
830 Stewart drive, #106,
Sunnyvale, CA 94085
Sunnyvale, CA 94085
