How High Peak developed an intelligence augmentation system for narrative building for the legal tech
Radhika Madhavan
Director of Marketing
Table of Contents
- The basics of intelligence augmentation
- Why do lawyers need an intelligence augmentation system?
- How did High Peak Software create a Matter Intelligence Augmentation System?
- Final thoughts
The basics of intelligence augmentation
Intelligence Augmentation (IA) combines the power of artificial intelligence (AI) with the perspective of human intelligence.
We mean that it helps humans make data-driven decisions—in a way, they would make it—and not a machine. IA focuses on augmenting intelligence to supplement the human-led decision-making process.
The difference between artificial intelligence & intelligence augmentation
While artificial intelligence and intelligence augmentation are used synonymously, there’s a key difference between them. AI refers to algorithmic models that mimic human intelligence to a certain degree and make decisions based on the structure of the model and input data.
IA, on the other hand, merely supplements the decision-making process. It analyzes and presents information in the user’s preferred manner and leaves it up to them to decide how to use it or what to do with it.
Simply put, you can refer to AI as the technology that focuses on doing activities that we generally consider intelligent/ smart. On the other hand, you can use the term IA when you’re referring to an application of AI. For example, if you’re talking about the model you’re using for data analysis, you can refer to it as an AI model—but if you’re talking about making decisions based on it, you can call the process IA.
Why should you use intelligence augmentation systems?
IA proves to be a helpful resource in a world lacking nuance as it helps businesses and their stakeholders make informed decisions.
A recent Gartner report indicated that 40% of I&O teams in enterprises would use augmented intelligence by 2023 to drive scalability and agility.
An excellent use case is to input existing data into an IA system and let it create in-depth insights on your preferred subject. This article will discuss one such use case in the legal industry and how they can leverage it to raise the bar for themselves, literally.
Why do lawyers need an intelligence augmentation system?
Lawyers work to win—but how do you do that when you don’t have the right systems?
Law firms are inundated with approximately 2000 to 5000 mailers a day. It makes it challenging to keep up with the massive onslaught of information. Moreover, it becomes even harder to store, process, build accurate narratives, and present that information meaningfully.
According to our client, 44th Street Technologies, their lawyers spend between 4 to 6 hours a day processing this information. Not only that, there are other challenges to consider too. Some of them include:
1. Excessive documentation
2000 to 5000 mailers a day is a lot to process. The sheer number of hard copies lawyers have to process poses a significant challenge. Because of the volume, critical documents go missing or get lost in transit or storage.
They also have to account for storage space to archive these documents for years once the case is closed. Hiring the right people to get this job done can be pretty time-consuming—making it an unnecessary hassle.
2. Lack of a consolidated system
Companies lack a proper documentation system, costing them a lot of money and losing cases over time. Without having an organized system to store all this information—it’s hard to know what’s relevant to the case and what’s not.
Also, it becomes complicated to understand the timelines of the event without having all the data in one place.
“During my time as a paralegal at a large law firm, we primarily used in-house software to manage, review, and prioritize documents for discovery productions, although I have experience with other e-discovery software available in the market.
The in-house software did little more than open the document for review. We were expected to analyze and prioritize the documents based on issues and relevant facts with separate word documents and charts.”
Joey Zawacki, Paralegal, 44th Street Technologies
3. Inability to build narratives
When lawyers don’t have all the information, it’s hard to pick out facts that help them win the case. It’s why it’s essential to record all the incoming data and use that information to tell stories that matter. It becomes tricky to build an accurate narrative of the case timeline if you don’t.
The most compelling court wins have always been built around a story based on hard-hitting facts. What’s a story without facts? Conjecture. That’s not what you want to be known for.
4. Too much time spent on analysis
Lawyers spend days and nights piecing snippets of the puzzle together. Without doing that, they don’t stand a chance in court. It’s one of the reasons they spend a lot of time analyzing the data they have.
In addition to that, they’re also trying to understand how they can use it. But with automation, you can reduce that time by a significant amount—and by that, we mean hours and not days.
5. Rising overhead costs
Law firms are, at the end of the day, a business. Every business wants to cut its costs as much as possible.
Law firms spend too much time and resources on storage space, junior employees (who waste time on redundant processes), legally compliant management systems, and more.
Over time, these add up, with no little to no returns in terms of investment. It’s one of the reasons we recommend digitizing your processes as soon as possible.
6. Security concerns
As a legal entity, any document that comes in and goes out is highly confidential. It means that law firms are always worried about whether or not their current systems are vulnerable.
While complete protection might not be possible, protecting it to the best of your abilities is in your best interests. We recommend using digital records because different records can be stored under varying levels of security. You can set the access levels based on who needs to access it.
Plus, they can always be backed up to cloud-based servers to ensure that they’re never lost. Paper records can easily get destroyed in fires or other disastrous events, whereas digital records can’t.
How did High Peak Software create a Matter Intelligence Augmentation System?
Given the challenges listed above, there are several objectives that 44th Street Technologies was looking to solve. Here are a few of them:
- Speed up analysis while accounting for every aspect of the case
- Capture notes while building narratives & timelines
- Create a consolidated document management system
- Access insights supported by visualizations after the analysis is done
- Create a digital archive of all cases—for future use
The technical team at High Peak Software employed several technologies to solve these challenges to solve these issues. Let’s have a look.
Automated document classification
The lack of a system to categorize unstructured data is a significant problem for document storage. Documents come in different forms: PDFs, images, chat logs, emails, videos, and other text-based formats.
While manual categorization is subjective—to circumvent this issue, we’ve used Machine Learning (ML) and Deep Learning (DL) models to categorize these documents. When the user uploads any document, these models read what kind of information and format they’re in and categorize them based on:
- Format
- Type
- Folder structure
It’s a simple and scalable solution that ensures lawyers are not wasting their time on documentation but on analysis instead.
AI-powered data extraction
When information is stored in the form of a hard copy, it’s hard to extract valuable snippets easily. Plus, the issue of potential damage to these records is always a glaring possibility.
Our engineers created an automated data extraction process that leverages Optical Character Recognition (OCR) technology to resolve this. These AI-powered models use data from handwritten text and annotated actual text values to understand what it needs to analyze. This capability can identify text when uploaded from a hard copy.
Now, lawyers can upload an image of a hard copy document, and the system will convert it into a machine-readable format. In addition to this capability, the data is categorized using deep learning algorithms and can be searched using the ‘search’ function later.
Document summarization
Nobody likes reading a heavily worded email to find the details that matter the most. Our engineers employed ML and Natural Language Processing (NLP) algorithms to auto-summarize emails.
Using this feature, lawyers can view short email snippets in the form of a thread—making them more digestible. It helps them stay focused on the bigger picture and spend time on other aspects that need their attention. This function also allows them to share it with other lawyers on the case—ensuring that they’re all on the same page.
Creating contextual guides
Once the data is uploaded onto the system, the NLP algorithms get to work. They categorize your documents into different Topic Clusters based on the details they contain. This grouping technology helps keep relevant case files in one place—and makes them easier to access whenever you need them.
Lawyers and paralegals can see what topics these documents fall under based on their case files’ dominant keywords and metadata. It saves them a lot of time when they have to find relevant documents.
Moreover, the algorithm also understands the emotions behind the text using Sentiment Analysis and classifies them as: positive, negative, and neutral. Given the nuances of their work, it fits in perfectly as it helps them understand the context and emotions behind their case’s data.
These details are in the form of text data which details several emotional states, contextual information about the case, and the people involved.
Knowledge management system
It’s essential to create a consolidated system—which is why it was our top priority. But, in this case, our technical team took it a little further. They created a knowledge management system that effectively manages all the content, indexes it for search, and organizes it based on a single parameter—the topic.
It consists of 6 categories:
- Facts: You can store any case-related insights, conclusions, and strategic pointers under this category.
- Notes: You can use this category to create highlights on specific facts that are already uploaded. You can also attach any reference notes if you have any.
- Issues: If there are specific issues you’ve noticed and want to record, you can do that and categorize them. You can add further context based on whether it is a high-level or incident-level issue.
- Source: Based on the source of your document, you can categorize and save them. For example, you can organize it under unprocessed files, processed files, email threads, audio/ video recordings, and more.
- Screenshot: This feature allows you to store screenshots of any documents for future reference. It’s beneficial when you have to reference a particular fact for the case.
- Address consolidation: AI models are powerful, but sometimes they might fall short in interpretation. If any errors occur in translation using OCR, you can correct them. For example, remove duplicate email IDs, etc.
Powerful data visualization
The main benefit of using artificial intelligence and intelligence augmentation in law is that you can create powerful storylines without analyzing them manually. The models do the heavy lifting for you, while your legal team can focus on more pressing tasks—working on the case.
Here are a few features that help you build strong storylines:
- Heat maps: You can use the heatmap feature to view events around a particular date for emails and other related documents. You can view it in a timeline format—making it easy to understand what happened and when.
- Email mapping: You can view the email exchanges between users using circos charts. You can filter the data based on subject, timeline, sender, recipient, etc.
- Timelines: Whenever you upload a document, the time and data are captured at the entry point. The system creates a timeline of events—helping you visualize the chronology of the case’s events.
- Email thread analysis: Since the system records all the data of each email, such as email, name, domain, source, etc., you can create a thread. It helps you determine which exchange needs to be attributed to who—keeping your factual information in check.
- Chronology: While it sounds similar to the timeline feature, the key difference is that you can view the processed data like events and snippets. You can export it for future reference too.
- Narratives: This feature ties everything together—it streamlines all the information in one place. Based on that, you can create an in-depth storyline and understand the actual context of your case files—without missing anything.
Final thoughts
Intelligence Augmentation, at its root, is built to help individuals make critical business decisions with accurate insights in hand. It can help you leave arduous manual tasks behind and focus on what matters—the data and the story it tells.
At High Peak Software, we’ve always been a huge proponent of creating software that makes a real difference for our clients. For 44th Street Technologies, we were able to build a system based on RPA and other technologies to help them consolidate, streamline, and supercharge their current workflows.
“The highlight of the solution would definitely be its ability to analyze relationships & ties within matter-related documents and provide users with its interactive visualizations. Consequently, giving end-users the flexibility to transform their notes into a story while doing any research or analysis.”
John Siniawski, CEO, 44th Street Technologies
If you’re looking for a custom-built software product for your business, reach out to our experts at High Peak Software today.
