Kyrylo Balalin

About the program

A hands-on live workshop for engineers, tech leads and CTOs demonstrating a complete engineering workflow with AI agents: from unloading an idea to verified implementation. Not code generation in a chat window, but a controlled delivery process. Format: two days, nine blocks. Day one covers theory: context mechanics, configuring a repository for agents, sub-agents and agent teams, the Ralph loop and model switching. Day two is practice: a full spec-driven development cycle live on a real product, from specify and clarify through design, tasks, implementation and verification gates. Total volume: 6 hours of material and 3.5 hours of live questions and answers. Participants received access to the recording, slides, code, skills, issue examples and the ralph-once.sh script.

Instruction language: Ukrainian

Learn more about the workshop →

Skills

• Artificial Intelligence (AI)
• Prompt Engineering
• Claude Code
• AI Agents
• Software Development
• Test-Driven Development (TDD)
• Software Architecture

Curriculum

1. 01Day 1 - Theory

   • Context as a resource: lost in the middle, context rot and why agents degrade
   • CLAUDE.md, rules and AI setup: repository config, path directives, four levels of rules
   • Sub-agents and agent teams: clean context, roles, shared task list
   • Ralph loop and model switching: advisor pattern, OpusPlan, verification

2. 02Day 2 - Practice

   • Specify and clarify: unloading the idea, PRD, acceptance criteria
   • Design: SAD, ADR, C4 - architecture decisions live
   • Tasks and test plan: decomposition, dependency graph, the basis for TDD
   • Implementation: vertical slices, parallel agents, verification gates
   • Q&A session: open questions and answers