Transforming How Lawyers Work: AI-Enabled Document Management

What happens when RAVN, the legal industry’s leading search and artificial intelligence (AI) platform combines with iManage, the market leader in document management (DM)?

Unlike the many specialized AI applications appearing in the legal market, this combination marries AI and DM into a single platform that we believe brings us closer to realizing a coherent user experience for lawyers in their practice of law. AI makes DM “smarter” and causes barriers to DM adoption to fall away. DM moves from simple document storage and sharing to supporting workflows tailored to the roles a person plays in a law firm, and AI delivers realtime, contextual information and tools that improve the speed, quality, and efficiency of legal work.


Fireman & Company was RAVN’s first American search partner, and we are currently leading the first two implementations of RAVN Connect (now iManage Insight) in the US legal market. We are also a long-time iManage partner. These relationships, combined with our deep insight into how lawyers work based on thousands of lawyer interviews we have conducted in recent years, provide us with a unique perspective on how these combined technologies will improve the practice of law.


Three key trends converged in 2017: The emergence of AI as the “next big thing” with immediately practical uses; the evolution of search from flat, two-dimensional look-up to intelligent, three-dimensional suggestions; and the transition of document management into a practice platform.

AI has become the “hot topic” for law firms. AI has been around for years, but interest has exploded in the past 18 months. In large part, this is because AI has become more practically achievable due to advances in the AI technologies themselves and dramatically reduced costs for the computing power needed to run them. Today we see AI embedded in specialist products for document review, expert systems, and contract analytics, and new products seem to appear almost daily. As a result, we see management at firms large and small discussing AI intensively as they seek opportunities for AI to improve efficiencies and lower the cost of legal services. We believe firms must look beyond the AI excitement and judiciously implement solutions that deliver real, measurable benefits and are adopted by lawyers. While there is value in point solutions targeted at specific business problems and use cases, there is even more value in AI platform solutions, like the iManage RAVN combination, that can be used to satisfy a broad range of use cases with a lower total cost of ownership.

At the same time, enterprise search has finally evolved from 2D to 3D. We consider firstgeneration enterprise search to be two-dimensional (2D) because it is flat: searches are informed only by the data sources indexed, the user’s search term, and limited inferences about relevance. Today, we are working with three-dimensional (3D) search, the next big evolution. Modern search tools such as RAVN are situationally aware. They make search smarter by taking context into account: the user’s past searches, other users’ searches, the searcher’s position or role in the organization, document histories, the context of a matter, and more. For lawyers, this means even more relevant and richer search results. As important, search features are embedded in and constantly running in other applications. That means Intranets, Microsoft Word, and document management systems can suggest to lawyers what content might be useful or what action to take next. Instead of lawyers going to search, search comes to them while they work (by running and analysing information behind the scenes), and with much better results and suggestions.

Finally, document management has transformed from an application to a platform. Historically, lawyers have been lukewarm about DM. We frequently see low adoption and avoidance behavior. This is not just inefficient; it creates a significant security risk for the law firm where client content is stored in unsecured locations. With next-generation DM, life is much easier for lawyers. Saving and retrieving both documents and email is simpler and integrated int0 workflows. Combining search, AI and DM now allows us to offer lawyers a practice-centric experience that is about more than saving and finding content. Practice-centricity seeks to understand why a lawyer is interacting with content (Reviewing? Drafting? Gauging status? Researching?) and suggests user journeys – a set of specific next steps – that can point the lawyer in the best direction and provide access to the most relevant resources inside and outside the firm. This, then, is where the AI component becomes the power behind practice-centricity.


iManage’s acquisition of RAVN will embed artificial intelligence as a core part of the document management platform. In our view, integrating flexible AI into DM that provides robust information governance and security features will lead to an even more powerful set of tools to support legal practice. At last, we see all the right tools come together to allow realizing our full practice-centric vision. As illustrated in Figure 1 below, AI-enabled DM will promote adoption and improve lawyer efficiency and effectiveness. Initially, it will remove past obstacles to using the DM by making filing, retrieving, and other basic operations simpler and easier. Beyond that, AI will be user aware and will be able to personalize and contextualize the individual user experience dynamically. This means that the system will be able to show a user related documents and email automatically based on the user’s history, the history of similarly situated lawyers, and the matter to which the document or email belongs. This could occur within a DM search or in other integrated applications such as the Intranet. Even more importantly, AI will deliver social network and knowledge graphs that show relationships among people and documents. This will help lawyers analyze how their work fits with prior work and tap into experts and related documents. And finally, over time, machine learning will be able to anticipate what a lawyer is working on from cues in the matter and what he or she types. From that, it will make suggestions about how to proceed, saving valuable time and improving quality and consistency.


Our work with our clients has proven that one of the biggest obstacles to lawyer adoption of the DMS is rigid file folder structures that don’t align with the way a lawyer or matter team prefers to organize a matter. To overcome this barrier, we have developed our proprietary design concept known as “organic matter management.” This enables a flexible folder structure that matter teams can tailor to their needs while also maintaining a core of consistent folders required by a firm’s records management policies and observing a firm’s information governance mandates.

Now, combine RAVN search and machine learning with Work 10’s flexible user interface and organic matter management moves to the next level.

  • Search recognizes who you are and your role on the matter team. Each lawyer can have access to a personal, working file view of the matter. Complex matters be simplified based on practice group designations; for example, a tax attorney can call up a matter file view focused on tax content, with key client documents alongside work by herself and the tax team.
  • Machine learning identifies and recommends best practices – within practice groups, specific teams and even specific lawyer combinations. This leads to the organic emergence of standards.

The end result is a user experience that is personal, practical and – most importantly – a major improvement on lawyers’ current workarounds.


Based on both rules and inferences, an AI + DM combination develops intelligence about what types of information to display for each role a user plays within a matter. Partners, associates, paralegals, secretaries, and staff have different needs and expectations when they save, search, and interact with information. Why should each member of a firm use a generic system? The answer is: They need not. We can take the role-based improvements that contribute to organic matter management, and further leverage this information by drawing inferences about the resource needs of each position. Some
examples illustrate this point:

  1. An associate doing substantive work should have access to related work product, matter context and knowledge content. She might also benefit from
    knowing which of her peers (within a couple of years of her bar admission) have done similar work so she can seek them out for advice.
  2. A partner needs to review the associate’s draft document. What is the status of the document and when were changes most recently made?
  3. An associate preparing to draft a deal document is able see how long others have taken to draft similar documents in other matters, so he can plan his time accordingly.


Incorporating AI into DM will bring improvements to KM, pricing, and legal project management (LPM) by expanding capabilities to extract and classify information, and combining information from multiple sources. Delivering actionable information to users within the context of the DM provides lawyers with a more compelling experience that will help promote adoption. These features will improve knowledge management (KM), pricing, and legal project
management (LPM).

Key Improvements for KM

On the deal side, AI-enabled DM search will be able to go beyond documents to clauses. As a lawyer is working on a specific clause, the system will be able to display comparable clauses from other documents, illustrating what’s market within the firm. It will be able to provide statistics on how many other documents (or matters) contain that clause type and pull available metadata from the matters for display alongside the word processor. When the lawyer saves a deal document, the system will be able to display an alert that the document contains atypical clauses for or missing clauses from the particular kind of deal document. This will create competitive advantages based on analysis of the firm’s extensive data collections, a significant leap forward for larger and merged firms.

For litigation, as a lawyer drafts a brief or pleading, the system will be able to auto-identify colleagues who have appeared in a jurisdiction or before a judge and identify similar language from comparable motions across cases – as well as whether the comparable brief or pleading succeeded or not. This will enable litigators to access work product more effectively and tailor arguments to specific judges more tightly.

Key Improvements for Pricing and LPM

AI-enhanced DM will help budget-managers find comparable matters based on the documents involved, and then pull the financials from those matters. Matter responsible partners will be able to compare burn rates to work status, based on document activity heat maps. We consider this to be a significant driver of LPM efficiency, while simultaneously reducing cost overruns and write-offs. Once a matter is underway, a lawyer or other professional monitoring progress against budget will have more than just billable hours compared to budget. Because the DM tracks exquisite detail about documents, including time spent drafting and editing, the system will be able to provide directional information about whether the hours remaining suffice to complete work at or under budget. These estimates of time required to complete work will close a major gap in many current LPM approaches.


As we stated earlier, until now search has not fulfilled its complete potential from the user’s perspective. It has been 2D and static. Users go to it. Results are deterministic. No longer. AI-enhanced search runs constantly in the background. It observes work as it happens and responds to the user’s position, role, and action as well as the nature of the matter. That means search becomes situationally aware and deeply informative, bringing to the foreground the firm’s resources most likely to be useful at each instance. With this three-dimensional, dynamic behavior, search moves to being an intelligent advisor, always at the ready. For example, the following functionality will be within reach:

  1. In addition to showing search results at the individual item level, the system can lead users to relevant clusters of people, documents, and other firm content. Effective user interface designs will allow users to easily change the level of detail visible for each cluster. In practice, this means that lawyers will virtually navigate the firm’s knowledge, resources, people, and experience the way they might have moved from one colleague’s office to her collections to a subsequently recommended matter, etc. We are, effectively, reconnecting the firm.
  2. The system can anticipate a user’s needs during a specific task by “knowing” something about the user’s task and automatically presenting additional information relevant to that task.
  3. If the user elects to use the suggested resources, clicking through leads to dynamic visual displays such as knowledge graphs or network diagrams. These displays allow finding the most relevant resource more quickly because they provide much more context than simple, linear hit lists. We can guide the user’s journey through clusters of people and data based on context and connections, which will in turn lead to more clusters as the user narrows her focus. Over time, as lawyers click through to see resources, the system will learn from those click throughs and improve suggestions.

In the past, it has often taken either very hard work or serendipity to assemble substantively helpful prior work and be informed about budget performance. At the risk of making an oxymoronic statement, welcome to planned serendipity.


Embedding AI in DM will make DM “smarter” and cause barriers to DM adoption to fall away. Lawyers will experience a more personalized form of organic matter management and the system will enable new levels of practice-centricity. In short, DM will work the way lawyers want to work.

AI functionality, working transparently behind the scenes, will surface useful information – people, content and connections – when and where it is needed based on work that a lawyer is performing and her role on the matter team. Moreover, this is the path firms must follow to improve value for clients. These improvements will increase lawyer efficiency, thus reducing costs for clients. At the same time, firms will benefit by reducing write-offs and protecting margins in fixed fee matters. Without question, the speed of technology is catching up to the needs of law firms and lawyers. The path forward requires hard work, but, with AIenabled DM, that path is visible and attainable.

For more information:

Joshua Fireman

Sally Gonzalez