Programming with AI Agents: Tips and Lessons Learned #263
Description
Title: Programming with AI Agents: Tips and Lessons Learned
Speaker: Sergei Makarov
Preferred month: Oktober/November
Duration: regular (30 min)
Language: English
Short description
In this talk, I will share my personal journey of building real-world projects with the help of AI agents such as Cursor, JetBrains AI, and GitHub Copilot.
I’ll present concrete tips for using AI effectively: how to structure prompts, avoid common pitfalls, manage costs, and ensure maintainability.
The session will also highlight the role of new protocols (like MCP) and tools (like Context-7) in making AI agents more reliable for everyday work.
By the end, you’ll have actionable strategies for turning AI assistants from “unpredictable helpers” into true productivity boosters.
Audience
This talk is aimed at software engineers, especially those curious about practical AI applications in development.
It will be useful for developers onboarding to new projects, engineers experimenting with AI-assisted coding, and anyone looking to increase efficiency while avoiding common mistakes.
Both beginners and experienced users of ChatGPT, Copilot, or similar tools will find valuable insights.