LangSmith Enhances User Experience with Homepage Redesign and Resource Tags
Iris Coleman Nov 19, 2024 11:58
LangSmith introduces a redesigned homepage and Resource Tags, improving organization for developers working on LLM applications, according to LangChainAI.
LangSmith has rolled out a significant update to its platform, unveiling a redesigned product homepage and the introduction of Resource Tags for enhanced resource organization, according to LangChainAI. These changes are aimed at improving user experience for developers working with large language model (LLM) applications.
Redesigned Homepage
The updated LangSmith homepage is now divided into three distinct sections: Observability, Evaluation, and Prompt Engineering. This new structure reflects the main areas developers focus on when building LLM applications. By organizing the homepage into these categories, LangSmith aims to streamline the developer workflow, making it easier for users to access the tools and information they need.
Observability
The Observability section includes features like Tracing Projects and Dashboards. Tracing Projects allow users to track and filter traces from their AI applications, providing detailed insights into application performance. Dashboards enable users to create custom monitoring charts to track important metrics over time, such as cost, latency, and quality, allowing for comprehensive performance analysis.
Evaluation
LangSmith's Evaluation section features Datasets, Experiments, and Annotation Queues, facilitating high-quality, continuous evaluations. Developers can curate datasets, run experiments, and score results with automatic evaluators. Annotation Queues help manage and organize data points requiring human review, ensuring consistent feedback and streamlining the evaluation process.
Prompt Engineering
The Prompt Engineering section includes tools for storing and versioning prompts, enabling collaboration and tracking of prompt evolution. The Playground feature allows for real-time iteration and side-by-side comparisons of different prompt and model combinations, aiding in the development of responsive LLM applications.
Resource Tags
To further assist developers, LangSmith has introduced Resource Tags, enhancing resource grouping and accessibility. These tags allow users to organize resources by custom criteria, such as environment or application, facilitating better management across different projects and teams. This feature provides flexibility in organizing resources, addressing the diverse needs of growing teams.
In future updates, LangSmith plans to implement Attribute-Based Access Control (ABAC) for Resource Tags, offering fine-grained access control based on tag attributes. This will enable secure and efficient resource management across various environments.
Image source: Shutterstock.jpg)