LeMUR is a powerful framework that allows you to apply large language models (LLMs) to spoken data, such as meetings, phone calls, videos, or podcasts, in just a few lines of code. One of the key features of LeMUR is its ability to generate concise summaries of your meetings.

To generate a summary, you first need to create a transcript of the meeting. LeMUR's Python SDK makes this process straightforward. Once you have the transcript, you can call the summarization endpoint, providing optional context information and specifying the desired output format (e.g., topic header and summary).

The generated summary provides a high-level overview of the key topics discussed during the meeting, including details such as metrics, issues, and other relevant information. This can be particularly useful for quickly understanding the main takeaways from a lengthy meeting without having to review the entire transcript.

In addition to generating summaries, LeMUR also allows you to ask questions about the meeting content and retrieve relevant answers. This can be helpful for quickly finding specific information, such as top-level KPIs or the number of days since the data team has received updated metrics.

Another useful feature of LeMUR is its ability to extract action items from the meeting transcript. This can help ensure that all the necessary follow-up tasks are identified and tracked, improving the overall productivity and accountability of your team.

LeMUR is designed to be highly flexible, allowing you to submit multiple files at once and handle large inputs (up to 1 million tokens, which is roughly equivalent to 100 hours of audio data). This makes it a valuable tool for working with extensive lecture series or other long-form spoken content.

LeMUR's question and answer capabilities allow you to easily extract insights from your spoken data. You can specify a set of questions and LeMUR will provide the relevant answers from the transcript.

For example, you can ask questions like "What are the top level KPIs for engineering?" or "How many days has it been since the data team has gotten updated metrics?". LeMUR will analyze the transcript and provide the corresponding answers.

This feature is particularly useful for quickly understanding the key discussion points and action items from meetings, calls, or other audio recordings. By leveraging LeMUR's question and answer functionality, you can efficiently extract the most relevant information without having to manually review the entire transcript.

LeMUR, a flexible framework, empowers you to extract actionable insights from spoken data with just a few lines of code. By leveraging the power of large language models, LeMUR can analyze your meeting recordings, phone calls, videos, or podcasts, and provide a concise summary of the key action items discussed.

To get started, simply call the action_items endpoint and LeMUR will deliver a comprehensive list of relevant action items from your input data. This feature helps you stay on top of your team's tasks and ensures that no important to-dos slip through the cracks.

LeMUR's versatility extends beyond action item extraction. You can also use the framework to generate meeting summaries or answer questions about the content, providing a valuable tool for staying informed and aligned, even when you can't attend every discussion.

With LeMUR's ability to handle over 1 million tokens, equivalent to approximately 100 hours of audio data, you can analyze extensive collections of spoken content with ease. Whether you're working with a single meeting recording or an entire lecture series, LeMUR has you covered.

Unlock the power of LeMUR and streamline your workflow today. Explore the comprehensive documentation and get started in just a few minutes.

LeMUR is designed to be highly flexible, allowing you to submit a wide range of tasks and prompts to the system. The custom task endpoint supports inputs up to 1 million tokens, which translates to roughly 100 hours of audio data. This means you can use LeMUR to process large amounts of spoken content, such as entire lecture series or extensive meeting recordings.

Furthermore, you can submit multiple files at once, enabling you to efficiently process batches of audio data. This flexibility makes LeMUR a powerful tool for a variety of use cases, from summarizing meeting discussions to extracting action items and answering questions about the content.

To get started with LeMUR's flexible task handling, check out the documentation and the welcome Colab notebook, which provide step-by-step guidance on how to leverage the full capabilities of the framework.

Lemur is a powerful framework that allows you to apply large language models (LLMs) to spoken data in just a few lines of code. With Lemur, you can easily generate summaries, ask questions, and extract action items from meetings, phone calls, videos, or podcasts.

The key features of Lemur include:

• Transcript generation using the Python SDK
• Summarization with customizable context and output format
• Question-answering on the meeting content
• Extraction of action items from the transcript
• Flexible custom task endpoint for any prompt or task
• Support for large inputs up to 100 hours of audio data

Lemur is designed to be user-friendly and efficient, making it easy to get started and unlock the power of LLMs for your spoken data. Check out the Lemur documentation and the welcome Colab to get up and running in just a few minutes.