Louisa Crawford
May 15, 2025 05:34
Interprup 2025, the first AI agent conference of Langchain, has introduced new product launches and industry insights. The main themes include agent engineering and AI observation.
Interprup 2025, the inauguration of Langchain, recently concluded from San Francisco and attracted more than 800 participants worldwide. The event was used as a platform for industry leaders such as CISCO, UBER and LinkedIn, sharing experiences and insights in the field of AI agents. According to Langchain’s blog, the meeting emphasized that the importance and potential of AI agents is increasing in various sectors.
Keynote speech
During the open keynote, Langchain’s Harrison emphasized the emerging discipline of agent engineering similar to software engineering, machine learning and product management. The keynote speech emphasized the need for expertise in business workflow coding, prompt and understanding to effectively develop AI agents.
Another pivotal theme was the dependence on the LLM (Language Model) application for various models. The Langchain Package, which recorded more than 70 million downloads a month, gained popularity with Openai SDK by providing the company with flexibility and choice in model integration.
Langgraph, the agent orchestration framework of Langchain, has been introduced as a tool for improving the reliability of AI agents by providing comprehensive control of cognitive structure. Through this framework, developers can distinguish them in the agent orchestration field by managing workflow and information flow.
Product release
Langchain has published some new products during the meeting. Currently, the Langgraph platform, which is currently available, provides distribution and management solutions for long -term running state agents. It supports a variety of distribution options, including cloud, hybrid and self -hosting settings.
Open source, No-Code Agent Builder Open Agent Platform allows non-developers to select tools, customize prompts, and integrate data sources through user-friendly interfaces to create an AI agent.
The Langgraph Studio V2 has also been released to provide a local running IDE for visualizing and debugging agent interactions. This version introduces new features such as Trace Pull-Down and introduces the prompt update directly in the UI.
In addition, Langgraph Pre-Builts aims to simplify the process of building a common agent architecture, while the LANGSMITH Observability contains agent metrics to optimize the application performance.
Future
In the future, Langchain introduced LLM-AS-JUDGE, a technology that evaluates AI performance that requires discretionary discretion. This method includes correction and audit scores with human feedback to ensure reliability.
The meeting had an important stage for Langchain in the mission to make AI agents a ubiquitous. The company plans to hold interrupts every year to promote continuous innovation and collaboration within the AI community. For more information about the meeting, visit the Langchain blog.
Image Source: Shutter Stock