ИИ-агенты — вот что действительно изменит разработку. Пишем ИИ-агент на Python, LangChain и GigaChat

Description:

0:00 The video opens with a discussion on how AI can write code, but its real value lies beyond code generation. 1:05 The presenter explains what an AI agent is. 2:00 He describes the paradigm shift that has been in software development for 80 years. 4:45 A simple agent that understands ordinary language is demonstrated. 5:55 The agent is prompted with a complex task. 8:04 The process of setting up a development task for an AI agent is outlined. 10:19 The presenter acknowledges that this is partly promotional content for GigaChat. 12:50 The tutorial shows how to build an AI agent using Python, LangChain, and GigaChat. 25:14 The newly created agent is launched. 27:15 The presenter reflects on how this truly changes the development paradigm. 27:53 He emphasizes the need to push further. 34:13 A multi-agent system is created. 38:06 The presenter invites viewers to reflect on what has been achieved. 38:30 The question of whether LLMs will replace BPM engines is raised. 39:51 The agent provides a highly flexible language interface to access business logic. 43:35 The presenter speculates that there will be more AI agents than people on Earth. 44:22 The agent correctly handles any input files. 47:55 New capabilities are showcased.