deepset Debuts Quantifiable LLM Response Accuracy with Latest Capability

By neub9
2 Min Read

deepset, the company behind the popular Haystack open source framework for building NLP services, has introduced a groundbreaking addition to its cloud platform. This new feature provides insights into the precision and accuracy of LLM generative AI (GenAI) responses. Learn more about deepset Cloud.

One of the main challenges with LLM-based GenAI is the issue of hallucinations, which can lead to data being placed in the wrong context or completely fabricated. deepset Cloud’s new capability aims to combat this problem by offering a measure of how well the generated responses are grounded in the source data provided to the model.

The Groundedness Observability Dashboard, part of deepset Cloud, uncovers trend data regarding how well GenAI responses are grounded in the source documents. It offers a quantifiable score that reveals the accuracy and factuality of an LLM’s output, enabling developers to make appropriate configurations to their systems, models, and prompts to produce reliable responses.

Additionally, deepset Cloud’s Groundedness Observability Dashboard can help organizations identify the optimal hyperparameters for their retrieval steps and optimize the amount of data fed into the LLM, ultimately reducing costs.

The platform also introduces Source Reference Prediction, which adds academic-style citations to each generated answer, referencing back to its respective document where the information was sourced.

In terms of data privacy, deepset ensures customer data is secured by robust security standards, adhering to SOC 2 Type II requirements. For enhanced security, enterprises have the option to run deepset within their private cloud environment.

With these new features, deepset reaffirms its commitment to building a robust trust layer within GenAI apps, allowing for reliably creating trustworthy applications with large language models.

For more information about deepset Cloud and its latest features, please visit

Share This Article
Leave a comment

Leave a Reply

Your email address will not be published. Required fields are marked *