transcribe

LangChain Academy New Course: LangSmith Agent Builder

LangChain · 1m · transcribed 17d ago
More from LangChain Business

Section Insights

# 0:00

Introduction to LangSmith Agent Builder

What is the LangSmith Agent Builder?

The LangSmith Agent Builder allows anyone to create production-ready agents without coding, using a chat interface.

  • No coding is required to build agents.
  • Users can create agents simply by chatting.
  • The tool guides users from idea to deployment.
# 0:20

Challenges in AI Adoption

What challenges do teams face when adopting AI?

Teams struggle with writing effective prompts, steep learning curves of automation tools, and uncertainty about which tasks can be automated.

  • Effective prompt writing requires practice.
  • Automation tools can be complex to learn.
  • Not all tasks are suitable for automation.
# 0:41

Advantages of Using Agents

How do agents differ from traditional workflows?

Agents can reason, adapt, and decide on actions in real-time, outperforming rigid workflows.

  • Agents can determine the right steps independently.
  • They can delegate tasks to sub-agents.
  • Agents continue working until tasks are completed.
# 0:53

Continuous Improvement of Agents

How do agents improve over time?

Agents utilize memory and user feedback to enhance their performance continuously.

  • Agents learn from user interactions.
  • Feedback helps agents become more effective.
  • Memory allows agents to retain knowledge for future tasks.
# 0:15

Simplified Creation Experience

What is unique about the creation experience with Agent Builder?

Agent Builder eliminates the need for complex workflows, offering a straightforward chat-based creation process.

  • No need for if-this-then-that logic.
  • Creation is simplified to a chat interface.
  • Combines power with ease of use.
# 0:28

Course Overview

What will the Agent Builder Essentials course cover?

The course will teach users how to build no-code agents and automate tasks, including a practical example of creating an email agent.

  • The course focuses on practical applications.
  • Users will learn to automate time-consuming tasks.
  • Building an email agent will be a key example.

Transcript

0:00 We’re expanding who can build agents. With LangSmith Agent Builder, anyone can now create production-ready agents without writing code, simply by chatting. Agent Builder guides you from idea to deployment, crafting prompts, choosing tools, and even generating sub-agents. In our new LangChain Academy course, Agent Builder Essentials, you’ll learn how to build these no-code agents with chat. Many teams face the same challenge when adopting AI. People want to use AI to boost productivity, but writing effective prompts takes practice, automation tools often have steep learning curves, and it can be unclear which tasks can be automated reliably.

0:41 Agent Builder solves this. Instead of forcing users to design rigid, step-by-step workflows, it creates true agents. Agents reason, adapt, and decide what to do next in real time. This is why, for so many productivity use cases, agents outperform workflows. They can figure out the right steps on their own, even delegating to sub-agents. They can work until the task is complete, looping through tool calls as needed.

1:11 And they can improve with your feedback, using memory to get better over time. With Agent Builder, we skipped the if-this-then-that canvas entirely and went straight to a fully agentic system that combines significant power with the simplest creation experience: chat. In this course, we'll walk through how to build an email agent and show you how to use Agent Builder to automate time consuming tasks. I look forward to seeing you in the course.

Summary

LangSmith Agent Builder allows users to create production-ready AI agents without coding, simply through chat interactions. This tool simplifies the process of building agents by guiding users from concept to deployment, making AI more accessible and effective for productivity tasks.

- LangSmith Agent Builder enables no-code creation of AI agents via chat.
- The tool assists in crafting prompts, selecting tools, and generating sub-agents.
- A new course, Agent Builder Essentials, teaches users to build these agents.
- Agents can adapt and make decisions in real-time, outperforming traditional workflows.
- They can delegate tasks to sub-agents and continue working until tasks are completed.
- Agents improve over time through user feedback and memory capabilities.
- The system bypasses complex workflows for a more intuitive agent-based approach.
- The course includes practical examples, such as building an email agent for task automation.

Questions Answered