KBR
KBR

hi there

[Kaggle Gen AI] Day 3 ๊ณผ์ œ ์†Œ๊ฐœ - Generative AI Agents ๐Ÿš€

[Kaggle Gen AI] Day 3 ๊ณผ์ œ ์†Œ๊ฐœ - Generative AI Agents ๐Ÿš€

Day 3์—์„œ๋Š” ์ƒ์„ฑํ˜• AI ์—์ด์ „ํŠธ(Generative AI Agents)์— ๋Œ€ํ•ด ๋‹ค๋ฃฌ๋‹ค.
๋‹จ์ˆœํ•œ LLM ํ”„๋กฌํ”„ํŠธ๋ฅผ ๋„˜์–ด์„œ, ์—์ด์ „ํŠธ์˜ ๊ตฌ์„ฑ ์š”์†Œ, ์˜์‚ฌ๊ฒฐ์ • ํ๋ฆ„, ๊ทธ๋ฆฌ๊ณ  ๊ธฐ์กด ์‹œ์Šคํ…œ๊ณผ์˜ ์—ฐ๊ฒฐ ๋ฐฉ์‹์„ ์‹ค์Šต๊ณผ ํ•จ๊ป˜ ์ตํžˆ๋Š” ์œ ๋‹›์ด๋‹ค.

๊ธฐ๋ณธ ๊ณผ์ œ์™€ ์„ ํƒ ๊ณผ์ œ๋กœ ๋‚˜๋‰˜๋ฉฐ, ํŒŸ์บ์ŠคํŠธ/๋ฐฑ์„œ/์ฝ”๋“œ๋žฉ์ด ๋ชจ๋‘ ์ œ๊ณต๋œ๋‹ค.

๊ฐœ์ธ์ ์œผ๋กœ๋Š” ํŒŸ์บ์ŠคํŠธ๊ฐ€ ๋ฐฑ์„œ ๋‚ด์šฉ์„ ์•„์ฃผ ์ž˜ ์š”์•ฝํ•ด์ฃผ๊ณ  ์žˆ๊ณ , ์„ค๋ช… ํ๋ฆ„๋„ ๋งค๋„๋Ÿฌ์›Œ์„œ ํ˜ธ๋‹ค๋‹ฅ ๊ณต๋ถ€ํ•ด์•ผ ํ•œ๋‹ค๋ฉด ํŒŸ์บ์ŠคํŠธ๋ฅผ ๊ต‰์žฅํžˆ ์ถ”์ฒœ!



๐ŸŽ’ Todayโ€™s Assignments

  1. Complete Unit 3 โ€“ โ€œGenerative AI Agentsโ€
    ์œ ๋‹› 3 โ€“ โ€œ์ƒ์„ฑํ˜• AI ์—์ด์ „ํŠธโ€
    • Listen to theย summary podcast episodeย for this unit
      ์œ ๋‹› 3์˜ ์š”์•ฝ ํŒŸ์บ์ŠคํŠธ ๋“ฃ๊ธฐ
    • To complement the podcast, read the โ€œAgentsโ€ whitepaper
      ํŒŸ์บ์ŠคํŠธ๋ฅผ ๋ณด์™„ํ•˜๊ธฐ ์œ„ํ•ด, ๊ด€๋ จ ๋ฐฑ์„œ ์ฝ๊ธฐ
    • Complete these codelabs on Kaggle:
      Kaggle ์ฝ”๋“œ๋žฉ ์‹ค์Šต ์ง„ํ–‰ํ•˜๊ธฐ
  2. [Optional] Advanced 3b - โ€œAgents Companionโ€
    [์„ ํƒ๊ณผ์ œ] ๊ณ ๊ธ‰ Unit 3b โ€“ โ€œ์—์ด์ „ํŠธ ์‹ฌํ™”โ€
    • Listen to theย summary podcast episodeย for this unit
      ์š”์•ฝ ํŒŸ์บ์ŠคํŠธ ๋“ฃ๊ธฐ
    • Read the advanced โ€œAgents Companionโ€ whitepaper
      ๊ณ ๊ธ‰ ๋ฒ„์ „ โ€œ์—์ด์ „ํŠธ ์‹ฌํ™”โ€ ๋ฐฑ์„œ ์ฝ๊ธฐ



๐Ÿ’ก What Youโ€™ll Learn

Learn to build sophisticated AI agents by understanding their core components and the iterative development process. Youโ€™ll also learn more about advanced agentic architectures and approaches such as multi-agent systems, agent evaluation and more. The codelabs cover how to connect LLMs to existing systems and to the real world. Learn about function calling by giving SQL tools to a chatbot (including an example using Gemini 2.0โ€™sย Live API), and learn how to build a LangGraph agent that takes orders in a cafรฉ.

๐Ÿ“– Day 3์—์„œ๋Š” AI ์—์ด์ „ํŠธ๋ฅผ ๊ตฌ์„ฑํ•˜๋Š” ํ•ต์‹ฌ ์š”์†Œ์™€ ๋ฐ˜๋ณต์ ์ธ ๊ฐœ๋ฐœ ๊ณผ์ •(iterative development process)์„ ์ดํ•ดํ•œ๋‹ค.

๐Ÿค– ๊ทธ๋ฆฌ๊ณ  ๋‹ค์ค‘ ์—์ด์ „ํŠธ ์‹œ์Šคํ…œ(multi-agent systems), ์—์ด์ „ํŠธ ํ‰๊ฐ€(evaluation), ๊ณ ๊ธ‰ ๊ตฌ์กฐ ์„ค๊ณ„ ๋ฐฉ๋ฒ• ๋“ฑ๋„ ํ•จ๊ป˜ ๋ฐฐ์šด๋‹ค.

โš™๏ธ ์ฝ”๋“œ๋žฉ์—์„œ๋Š” LLM์„ ์‹ค์ œ ์‹œ์Šคํ…œ๊ณผ ์—ฐ๊ฒฐํ•˜๋Š” ๋ฒ•์„ ์‹ค์Šตํ•˜๋ฉด์„œ

  • SQL ๋„๊ตฌ๋ฅผ ์ฑ—๋ด‡์— ์—ฐ๊ฒฐํ•ด ํ•จ์ˆ˜ ํ˜ธ์ถœ(Function Calling)์„ ์‹คํ–‰ํ•˜๋Š” ๋ฐฉ๋ฒ•,
  • Gemini 2.0 API๋ฅผ ํ™œ์šฉํ•œ ์‹ค์‹œ๊ฐ„ ์˜ˆ์ œ,
  • ๊ทธ๋ฆฌ๊ณ  LangGraph๋กœ ์นดํŽ˜ ์ฃผ๋ฌธ์„ ์ฒ˜๋ฆฌํ•˜๋Š” ์—์ด์ „ํŠธ๊นŒ์ง€ ์ง์ ‘ ๋งŒ๋“ค์–ด๋ณธ๋‹ค.