iManage’s London ConnectLive conference last week saw some important observations and unexpected revelations, as CEO Neil Araujo kicked off the day by saying: “We’re working hard because we recognise that we’re at an important moment in history and because of the game changing impact AI will have on knowledge work.”
While AI was inevitably a recurring topic, Araujo introduced another recurring theme for ConnectLive, which is that AI is like electricity. “Only when you put it to specific use does it have meaningful impact,” he said. “Electricity is only useful when you apply it to darkness with a lightbulb and it enables you to see in the dark. Or it enables you to make toast. Or have heat. Similarly, AI only has value when you apply it to the outcomes you are seeking in the business. The first step is to think of those outcomes: where do you want to improve the efficiency or effectiveness of what you are producing, or improve quality of life? Outcomes are often the result of changing business processes. Identifying those processes is a critical piece.”
Araujo said that core to success is that you also need a strong foundation and AI is dependent on three things. The first is data. “Everyone will have access to the same large language models, what will differentiate it is how you apply your own unique content through processes, so that you get outcomes that reflect the special sauce you have,” he said. Second is data governance, and Araujo asked: “Why is that important? So you’re not sharing information that is confidential or in breach of regulation.” The third is having a cloud platform. “There is a fair amount of complexity for you to effectively use AI and it is very challenging to do that on premises,” Araujo said.
Fireside Chat
In a fireside chat with Araujo, Anne-Marie Ovin, CIO and head of knowledge, process and innovation at Swedish firm Vinge, and Linklaters’ CTO Bruna Pellicci, discussed their seemingly seamless moves to the cloud.
Pellicci joined Linklaters five and a half years ago when the firm’s Documentum system had been developed into ‘something of a monster’, with layers and layers of third-party plugins. Nonetheless, as these things go, it seemed that users loved it, Pellicci observed.
In the magic circle firm’s journey to the cloud, Pellicci said: “We sat down and created a roadmap of where we wanted to be with technology; not just the document management system, but what would make us innovative and leading edge. A key thing was the DMS. There was a big drive to move to the cloud and everything that goes with that, including better security.” Key was having “the right system to plug everything into” and also iManage’s partnership with Microsoft.
Linklaters – here comes something of a revelation – has recently gone live on iManage Cloud across Asia, Europe, the US and then UK, with what Pellicci described as “fantastic help” from Morae Global and Encore Technologies. After going live on the new system, Pellicci said: “It was so quiet, I thought ‘no-one is using it.’ Threat Manager shows where people are using it, and it lit up. It just went really well. It’s easy to use and now people have DMS on their mobile. The fact that we left a system that everyone loved to zero complaints is unbelievable.”
Linklaters – another revelation – has just recruited a new head of AI in the technology team and is rolling out Copilot, which was announced just a few days before the conference. Pellicci said: “We’ve rolled out Copilot and we’re looking at how can we harness that and what do we do around knowhow. We’re looking forward to working with iManage in the AI world.”
While not working at a firm of the size of Linklaters, Ovin said that at Vinge, the firm had previously had siloed on premises systems and wanted to move from focusing on maintaining a system to focusing on the content within it. Vinge, too, wanted a strategic technology partner, and having one that was close to Microsoft was key. “Everyone says they are a good partner of Microsoft, so we asked Microsoft who they are closest to, and they said iManage,” Ovin said.
In a great analogy, Ovin described swapping out a DMS to flying an airplane and changing the engine at the same time but says: “It was the smoothest project ever at the firm.”
Product Roadmap
Fireside chat complete, up next at ConnectLive was a fact-filled roadmap keynote from iManage’s EVP of product management, Shawn Misquitta. One of the key priorities for iManage, Misquitta said, is how to eliminate friction, “or the data won’t be in the right place.”
“Once the data is in, it’s what you’re doing to protect it,” he said.
A key tenet of Misquitta’s presentation was how to eliminate administrative tasks to make more time for higher level tasks such as advice. “What if you can eliminate mundane work and give people superpowers?” Misquitta asked.
One of the mundane tasks is email, where lawyers’ mailboxes are quite often overrun with messages. When emails come in, they need to be filed. “Now every morning, you get an email into your mailbox suggesting where things need to be filed. If you agree, it does it. It looks at attachments and suggests where to file them. You can save hours in a month by taking away these mundane tasks,” Misquitta said.
Another time-consuming task is deal closing. iManage Closing Folders has a check list to obtain all eSignatures and ensure the signatures are in the right place. “It’s an example of how we’ve taken signature detection modules to take away the drudgery and risk of assigning the wrong signature to the wrong thing,” Misquitta said.
Within the documents themselves, Misquitta said that iManage can classify documents by learning what is in the content. “You give a thumbs up, thumbs down so it learns,” he said. “These are custom AI modules our data science team built from scratch. You can’t use generative AI at scale, or the bill will be so high, but we have built it from scratch to do it at scale and we have built it into the fabric of the cloud, so the data never leaves.” This theme of using ‘old’ and ‘new’ AI to best effect is important and was revisited multiple times during the day.
Ask iManage is iManage’s gen AI offering where you can take a document or set of documents and ask questions either ad hoc or based on templates and Misquitta said: “We can’t expect knowledge works to be prompt engineers.” This, again, was revisited in a later AI session.
Copilot is the big question on everyone’s lips and Misquitta observed: “A lot of iManage customers ask, ‘Do I need one or both?’ You need both. The reason is that Copilot understands what is on the web and in the M365 ecosystem, but it doesn’t understand what is in iManage. In collaboration with Microsoft, we have built a plug-in so that Copilot understands how to search the DMS and fetch the right documents. You can then find the terms leveraging Ask iManage for a lot of these functions.”
In turning humans into superhumans, Misquitta said that there is a lot of specialised data and systems that iManage can’t replicate. “Instead the approach we’re using is to make sure our AI integrates with, for example, Vincent AI from vLex so knowledge workers can be productive wherever they are.”
New on the roadmap are iManage Tracker, iManage’s engagement centric task management solution, plus co-authoring. “Microsoft is working with us so we can have co-authoring across the whole ecosystem,” Misquitta said. Over the next few months you’ll be able to choose who you want to collaborate with and create a link with an expiry data to avoid the problems around creating and changing permissions.
AI Use Cases
Last but not least of the key sessions in the morning at ConnectLive, VP of product Jan Van Hoecke and legal practice lead Jenny Hotchin took to the stage to discuss AI specifically, although Van Hoecke said: “AI is a general purpose technology. Neil’s electricity analogy is good. I want to talk about how it can make our lives easier.”
Van Hoecke, who was a co-founder of RAVN (acquired by iManage in 2017) said that AI does or will touch all parts of our lives, from driving to how we get our information. The market has moved faster than anticipated and Hotchin, who was previously a lawyer at Pinsent Masons, said: “AI is about keeping users at the centre of use cases.”
Hotchin and Van Hoecke discussed the use case of fee-earner ‘Thomas’, who wants to find information from a document using Ask iManage, which enables users to find specific data points within documents, ask follow-up questions, and take specific action. In the example, given, instead of everyone doing something different, there are 10 template questions, which are also used as grounding data. Van Hoecke said: “Gen AI models make stuff up; that’s what they were trained to do. Grounding is not easy to do reliably and it’s key that the answers are grounded into the document it’s looking at and nothing else.”
Interestingly, there is a new feature that will turn the information from Thomas’s questions of multiple documents into a table, which could become part of an automated workflow.
When it came to iManage’s ability to leverage AI across the whole of the DMS, Alex Smith, global product lead (senior director) for knowledge, search and AI, appeared to gate crash the stage to add his thoughts on information architecture and enriched metadata. “We want to make it as simple as possible to curate knowledge,” he said. “Of course we give your users the option to provide metadata, but it’s not required. That’s where AI can kick in, with pre-filled metadata they can review, adjust and go-forward.”
Concluding Thoughts
Speaking to Legal IT Insider after the conference, Araujo said: “If you can get a robot to tag all your data with a high degree of accuracy and without compromising confidentiality, that is a huge leap not just for AI but for information management.
“There are two fundamental reasons that people use the DMS; one is version control and the second is recall. That tends to be an imperfect process because human memory by definition is flawed and sometimes you don’t know what to search for but if its tagged, you can get to something you didn’t remember existed.
“The approach we’re taking to AI is multi-prompt: Ask iManage and also traditional AI and ML in tagging, which means we can run across the entire DMS and 100s of millions of documents without incurring the cost that it would take using other advanced technology. It’s taken us years to build.”
He adds: “The key point I wanted people to take away from ConnectLive was that managing your data and governance are a pre-requisite for AI and when you want to work with AI, start with the outcome and processes.
“The comment that I got back from the conference is that ‘we now understand we should be looking at what should our roadmap be for using AI at scale, in production, in a way that will change the way people work.’ People are genuinely excited about that possibility. The chatter has gone from the fear of AI and losing jobs to how to use it and get end user benefit. That I think is the take away. The conversation is not about fantasy and science fiction, but how do we get real.”