AI-Native Development Practices for Modern Developers

Main Speaker

Learning Tracks

Course ID

42891

Date

21-06-2026

Time

Daily seminar
9:00-16:30

Location

Daniel Hotel, 60 Ramat Yam st. Herzliya

Overview

This seminar provides a practical introduction to modern AI coding assistants, with a focus on GitHub Copilot as the main tool used throughout the training. Participants will learn the core concepts behind LLM-based coding assistants, how to work with chat effectively, how to improve results through context and instructions, and how tools such as MCP expand what these systems can do inside real development workflows. Although the seminar is demonstrated with GitHub Copilot, the same core ideas apply to other coding agents and assistants as well.

Who Should Attend

Prerequisites

  • Basic familiarity with software development and working with code repositories
  • Access to any coding assistant or coding agent, such as GitHub Copilot, Claude, Cursor or a similar tool
 

Course Contents

Session 1: What is an LLM, what is an agent, and where Copilot fits
  • What an LLM is and how it generates text, code, and explanations
  • The difference between an LLM, a chatbot, and an agent
  • Strengths and limitations of LLMs in software development
  • Where GitHub Copilot fits in the AI assistant landscape
Session 2: Getting started with Copilot Chat
  • Using Copilot Chat to understand code and navigate a codebase
  • Asking Copilot to generate code, explain logic, and suggest improvements
  • Using Copilot for refactoring, debugging, and writing tests
  • Common prompt patterns for better results
Session 3: Prompting with context
  • Why context matters when working with Copilot
  • Using selected code, open files, and repository context effectively
  • Improving prompts with file references, chat variables, and workspace awareness
  • Comparing vague prompts with precise, context-rich prompts
Session 4: Custom instructions
  • What custom instructions are and why they matter?
  • The difference between personal, repository, and organization instructions
  • Using instructions to reflect coding standards, architecture, and team conventions
  • Writing instructions that improve consistency without over-constraining results
Session 5: Repository instructions in practice
  • Creating and maintaining.github/copilot-instructions.md
  • Translating team expectations into clear natural-language guidance
  • Using repository instructions to improve code style, tests, and documentation
  • Verifying that Copilot is applying repository-level guidance
Session 6: Using MCP for richer context
  • What MCP is and why it expands Copilot beyond the local repository
  • How external tools and systems can provide better context to Copilot
  • Practical examples of using MCP with repositories, issues, and pull requests
  • Security, control, and adoption considerations when using MCP
Session 7: Governance, usage, and spending
  • Understanding Copilot usage, entitlements, and premium requests
  • Tracking adoption and monitoring how Copilot is being used
  • Managing budgets, allowances, and spending controls
  • Building a responsible rollout strategy for teams and organizations
Session 8: Applying Copilot in real development flow
  • Using Copilot across the full development lifecycle
  • Combining chat, context, instructions, and external context effectively
  • Identifying strong use cases for teams and individual contributors
  • Defining best practices for daily engineering work

The conference starts in

Days
Hours
Minutes
Seconds