Using AI should be the careful application of YOUR knowledge and YOUR insight developed over your years of practice to more quickly, and perhaps for younger associates, more skillfully, provide solutions for your clients. Like in many systems, the issue is one of garbage in and therefore garbage out. The most skillful application of AI in iManage systems and most all others requires the curation of the data set to which AI-born questions are applied.
The question then is how to curate your data. Firms using iManage have several advantages in how to begin curating relevant data for inclusion in iManage AI directed data. Litigation firms can tag documents (iManage allows many different database fields to identify document) based on different doc types, full text searches, document dates and legal precedents, among others.
Once an AI repository has been created or tagged within iManage, users will be able to find good data, relevant to their issue using English-language (and a few others) inquiries. One goal could be to write new documents in a ‘voice’ common to the firm by incorporating curated concepts and phrases from previous relevant documents. Another goal could be to shorten the time required to write a new document, on point to the issues of the new client, but incorporating proven ideas, concepts and published legal precedent.
It is important for a firm to assign the initial data curation to a small team or even a knowledge manager to identify curated documents. Consistently applied rules used to identify AI candidate files will help initialize and later maintain a clean data set. The firm should develop an initial set of metadata terms / rules to apply to curated documents. Your data set should contain the most relevant and valuable information from your iManage system. Questions to ask yourself as you evaluate your data include:
- Is it well written?
- Does the document follow a clear, logical chain of thought?
- Is it relevant?
- Is it up to date?
- Does it follow legal best practices?
- If it was filed, does it represent legal precedent?
- Does it represent the opinions of the firm principals?
- Is the document an internal or external document – you can apply different weights to the AI results
Once you have identified documents for inclusion, we suggest tagging the documents with a discrete field to identify that document for AI use. You can also apply additional metadata at that time or as your AI repository grows to enhance searches and the application of AI to your client’s documents. These might be tagging a document with:
- Areas of law
- Practice area
- Additional conceptual phrases
Client and matter information is already extant as well as doc types since these are iManage documents.
Searching with AI enhancements will point out documents that fit the criteria of the search, plus will highlight the pertinent sections of the documents. The ability to highlight the portions of a document that are on point can save many hours of reading brief after brief. Using Westlaw or Lexis thereafter to check the legal precedent is faster and more accurate.
One important reason to use a small team to identify documents for inclusion to the AI repository is that they can more easily apply the clean metadata criteria already developed by the firm. This same team should also be charged with adding new information to the AI repository and removing stale data. An important concept as you develop the use of AI for your firm is that the initial metadata criteria should also be reviewed periodically as AI is inherently not written in stone. AI is a dynamic concept that can grow and develop new ideas as your firm grows and changes.
There are several iManage modules to include in moving to an AI-enhanced model for your firm. Contact an Innovative Account Executive today to learn more about how Innovative’s engineers can help you develop curation rules, curation metadata, tags, and conceptual phrases to identify AI related documents.
Contact us now to unlock the full potential of AI and iManage!