Early Access
This article has been published prior to Ask’s GA date. Content may be subject to updates before its official release.
With the rapid advancements in Generative AI and large language models, it has become clear that this technology can be applied to the eDiscovery space, enabling users to explore and analyze case documents in entirely new ways across all stages of the EDRM.
Ask brings this capability directly into LGK, allowing users to ask natural-language questions of their project data and receive AI-generated answers.
What is Ask?
Ask enables users to ask natural-language questions about their case data and receive AI-generated responses grounded in the documents available in their project (entire corpus or filtered subset).
Grounded in source documents with transparent citations and links to referenced material.
Augmented with related documents, showing up to 100 semantically similar items that may be relevant even if not directly cited.
Why It Matters?
Ask addresses key challenges in legal review and investigation workflows:
Time-consuming review: Legal teams spend significant time manually reading through large sets of documents to answer case-relevant questions.
Lack of trust in AI: Without transparency or citations, users struggle to trust AI-generated answers.
Missed context: Conventional keyword search can overlook semantically similar documents, leading to gaps in review.
Information fragmentation: Answers are often scattered across multiple documents, requiring manual synthesis.
Goals and Objectives
Ask is designed to help users work smarter and faster by:
Accelerate early case understanding by providing direct, cited answers to complex questions.
Build trust in AI output through transparent, document-level citations.
Expand discovery workflows with semantically relevant suggestions beyond keyword matches.
Reduce cognitive load by centralizing answer synthesis and supplemental evidence in a single interface.
Key Use Cases
Ask helps users answer complex questions and uncover relevant context quickly. Examples include:
Reviewer Question: “What did the executives say about the company’s financial risks in Q3 2021?” → Ask provides a synthesized answer from emails, with footnoted citations.
Litigator Validation: “Was there internal discussion around the XYZ acquisition before April?” → Ask delivers a yes/no style response with references, plus additional related documents.
Compliance Investigation: An investigator queries potential misconduct. → Ask generates a summary with citations and semantically aligned documents for deeper exploration.
Availability
Ask is being rolled out in LGK as an add-on feature. It can be enabled at the instance (MSA) level for now. For activation details, please contact your Customer Success or Sales representative.