By applying the latest AI trends and techniques, LexRock AI is leading the world in document comprehension and analysis

Until recently, the conventional wisdom was that while Artificial Intelligence (AI) was better than humans at data-driven decision-making tasks, it was still inferior to humans for cognitive and creative ones. But in the past two years, language-based AI has advanced by leaps and bounds, changing common notions of what this technology can do. The most visible advances have been in what’s called “natural language processing” (NLP), the branch of AI focused on how computers can process language and make computer languages more dynamic so they are easier for humans to understand. In its use, NLP technology turns out to have various capabilities that can help human life become more effective, especially in the business world.

LexRock AI is one such company, comprised of a creative team of AI experts, software engineers, and developers who create tools that facilitate the extraction of information from any documents through the use of AI (NLP, Machine Learning). Its solutions serve the insurance and financial sectors. It also helps customers comply with data protection legislation.

LexRock AI was founded in 2019 and is based in Montreal, Quebec. In February 2023, LexRock AI was selected by the Canada’s AI Global Innovation Cluster, Scale AI, a as part of a delegation of 12 Canadian companies and organizations to promote our national AI ecosystem at the World AI Cannes Festival (WAICF2023).

Alain Lavoie, CEO of LexRock AI, spoke exclusively to CIO Bulletin on how his company-built products are helping their clients’ increase productivity in a more quick and cost-effective manner.

Interview Highlights

Q. Give us a brief overview of LexRock AI’s past. Was there any prior experience that influenced the company’s formation?

The LexRock AI story began in mid-2015, at a time when artificial intelligence was beginning to attract a lot of interest. Our team had over 20 years of combined experience in autonomous natural language processing (NLP), information retrieval (IR), and information extraction (IE). We made the decision to use this knowledge and our intellectual property to benefit companies by providing cutting-edge solutions in the three growing industries of finance, insurance, and law.

Since 2015, we have invested over US$6 million in the development of a unique technology platform, following a trial and error process supported by constant technological advances. Today, in 2023, we are proud to have a great market fit platform that enables our customers around the world to solve everyday problems.LexRock AI is well-positioned to be a global leader in text and document analytics to increase operational effectiveness, revenue, and profitability for organizations because to our platform, 27+ years of NLP experience, and dedication to developing new solutions.

Q. What are the values of the company? What qualities do you seek in employees who will uphold those values?

Respect, inclusiveness and diversity, honesty, excellence, focus, and pleasure are among our company’s core principles. We value respect because it promotes team members’ ability to work well together. In order to create an agile team with a wide range of skills and experiences, diversity and inclusion are essential. Integrity is essential since it increases our credibility in the eyes of our customers. To excel, we must continually innovate and have a strategic mindset in order to be competitive in our market. Focus enables us to maintain alignment with both our clients’ needs and our vision and action plan. Pleasure is significant to us because we want our staff to enjoy their time at work.

To find employees who will uphold these values, we look for people who are passionate and aligned with our values. We value respect, honesty, integrity, innovation, attention to detail and customer focus. We also look for people who are determined, strategic and tactical, and who have a positive and collaborative attitude.

Q. LexRock AI initiates the development of platforms using AI to meet the needs of four industrial sectors (financial, legal, insurance, and regulatory compliance) in task-based tools for legal professionals and non-professionals. How do you interpret this further?

There is a lot of competition among the players in each of these sectors. They are all looking to differentiate themselves from their competitors through artificial intelligence.

The finance, insurance, and legal services industries are attracted to technologies like LexRock AI that use natural language processing (NLP) for several reasons.

First, technologies like ours can be used to assist employees with routine tasks such as document analysis and contract review, saving time and increasing operational efficiency.

Also, we can assist in enhancing data analysis and comprehension, which is essential in the financial and insurance industries where decisions are frequently dependent on accurate and trustworthy data. In order to identify trends, dangers, and opportunities, businesses can utilize LexRock AI Technology to extract information from massive amounts of unstructured data, like research reports, press announcements, and newspaper articles.

Q. What has been the biggest obstacle for your business? How did you deal with it?

We had a bold vision that presented us with some challenges. We wanted to create a flexible and agile platform that could quickly respond to targeted use cases for our clients. A platform that integrates with our clients mission systems (e.g., Fenergo, GuideWire) and always utilized the latest AI scientific advancements. A platform where it was possible to create an application that precisely met the needs of end-users in their daily tasks. In short, our platform solves real-world problems and improves the operational efficiency of businesses.

From a technical perspective, we also gave ourselves some challenges. We wanted to enable anyone to create an end-to-end artificial intelligence that met their needs without the help of an AI specialist. We also believed that models and algorithms would become commodities in the future, and we needed to be able to design an agile platform that could continuously integrate scientific AI advancements. For example, the release of ChatGPT was not negative news but rather a great opportunity to offer this advancement to our clients to solve their problems.

We also gave ourselves an additional challenge by choosing the PDF format as the pivot format. A PDF document is visually appealing but generally very difficult to manipulate from a computer science standpoint. It is one of the most widespread formats. We put a lot of effort into developing our own libraries to extract high-quality text from PDF documents; like high-performance cars, the higher the octane rating of the fuel, the more powerful the engine. Therefore, we made it our mission to increase the octane rating of the text extracted from PDFs in order to provide the best possible information to AI engines.

We set the bar high with a very bold product vision. But as they say, “If it were simple, it would be a common practice for everyone.” The main obstacle we faced was ensuring that we designed an agile and sustainable product while everything in AI was moving at breakneck speed. To achieve this, we made the following development assumptions:

• Unstructured data processing will be in high demand in the coming years.
• The data belongs to the customer.
• We must be able to annotate and extract information from a PDF document better than Adobe or Artifex.
• The platform must be able to quickly integrate any scientific AI advancements, whether from academia or the GAFAM.
• The platform must allow for the rapid construction of targeted and customized applications to meet real-world use cases.
• We cannot compete with the big players, who have enormous resources. Let’s partner with the best university researchers and major AI producers and integrate their AI models.
• The platform shall run in the cloud, in the customer’s cloud, or on the customer’s on-premise servers.

We are located in Montreal, Quebec, Canada, one of the world’s best AI ecosystems (Canada ranked 4th and Quebec 7th according to the Tortoise study), with a very high concentration of researchers,including Yoshua Bengio, one of the three godfathers of deep learning. We were fortunate to be very close to this ecosystem in order to have a clear view of the next AI advancements and thus be able to carve out our own niche.

We are recognized in Quebec as highly innovative and creative. This is what we have done to position ourselves where we are today with very modest means and an extraordinary team.

Q. Through the use of artificial intelligence, LexRock AI creates decision-support tools that draw attention to unstructured textual data in documents. How well do you address the concerns of your clients in light of this?

Manual data entry from documents is one of the most time-consuming, costly and error-prone tasks for an organization. Our customers are seeking a solution that will enable them to increase the effectiveness of their business operations. They will be able to raise their sales and profits as a result.

By providing a setting that is conducive to their everyday tasks, they will be able to get around the labour shortage. And that’s exactly what we provide—a platform fueled by AI that aids people in their daily activities.

Imagine that I offer you a very special pair of glasses. Ultra-specialized glasses that help you read documents very quickly and extract relevant and useful information for analysis For example, think:

• The financial analyst who has to read a financial statement and extract relevant information for calculating financial ratios;
• The actuary of an insurance company who must read a health insurance booklet to extract the terms and benefits in order to establish an insurance premium proposal;
• The investment fund manager, who must analyze a company’s ESG performance before investing;
• To the lawyer who needs to review or compare a set of clauses as part of a due diligence review.

You will agree that these tasks can be quite long and tedious, and that they often include transcription errors (copying from a PDF to an Excel file). In short, tasks that are too often boring and do not really have any added value.

Now imagine these same glasses adapting to the reading you do. Better yet, you can adjust them to your vision without the need for specialists. You can make your own adjustments to find information faster.

That’s exactly what the LexRock AI platform offers its users. We have succeeded in designing a platform that can understand and learn from humans and teach an AI how to do it. A platform that can search, locate, retrieve, and compare any type of data from any document. A platform that allows any professional to train their own artificial intelligence with their data without having any algorithmic notion of AI.

Our platform accelerates productivity, improves performance, and increases organizational efficiency while significantly reducing the cost of text and document processing. Some of our customers have saved up to 90% in time. What used to take hours now takes minutes.

Q. What new projects is your business now working on?

There are currently several economic uncertainties that impact many countries and industries. Geopolitical tensions, interest rates, and inflation are just a few examples. These economic uncertainties can have a significant impact on businesses, consumers, investors, and financial markets and can cause a slowdown in the launch of innovation projects. It is important to remember that an AI project is considered an innovation project. And when these major economic uncertainties arise, companies become more cautious and cut back on new IT projects, especially innovation projects.

This context challenged us from the beginning of 2022, so we decided to start two projects in sectors that should be much less impacted, i.e., regulatory compliance and cybersecurity. The first one, LR Protect, allows companies to anonymize PDF documents and thus comply with privacy laws and regulations. The second project, “Dark Octopus”, which is scheduled for release in late 2023, will enable cybersecurity specialists to navigate the DarkWeb for cyber threats. We are designing glasses that will allow specialists to detect information on the DarkWeb more quickly.

With the help of these two technologies, LexRock AI will be able to expand into new markets and instantly change course in the case of a recession or downturn.

Q. Where do you anticipate your business to be in a few years?

The NLP market is growing rapidly. According to Priority Research, the global natural language processing market size is predicted to expand to around USD 361.6 billion by 2030 and grow at a CAGR of 39.06% from 2022 to 2030. The markets we are targeting, i.e., finance ($20.5 trillion), insurance ($6.1 trillion), and legal ($668 billion), are also huge.

Our platform, in our opinion, has a lot of promise. Our flexibility and capacity to effectively integrate the most recent AI algorithms published, among other things, allow us to develop in harmony with the top companies in our field (GAFAM).

We have built an agile, flexible, multilingual and scalable platform where the human is always in the loop and where the artificial intelligence adjusts according to the human’s behavior. This guarantees our customers that they can grow with us.  

Our vision is to become a world leader in intelligent document analysis based on our high technological agility. That’s where we’ll see us very soon.

Tell us about your team.

I had a dream in my early 50s to create a viable start-up with rapid expansion, and that goal became LexRock AI. Thus, for this fantastic trip, I surrounded myself with young and experienced people. I wanted to demonstrate that a blend of youth and grey hair might achieve tremendous things.

All of the members of my management team are in their 30s. They are all highly brilliant individuals who want to see this business expand. My successors, who will take the helm and advance this business, are being trained by me.

The success of a company is not the story of one person, but rather of a team. And at LexRock AI, we have a team of brilliant people who care about the success of the company. It’s really nice to see them all working together. We also have a great advisory board and a lot of people helping us grow.

Thank you to everyone who helps us! Thank you to my entire team! You are the best. I am truly fortunate.

About | Alain Lavoie

Alain Lavoie is the Chief Executive Officer of LexRock AI. He has been working in the field of information management and natural language processing (NLP) for more than 25 years, developing tools, algorithms, and products based on NLP and modern machine learning and artificial intelligence (AI) techniques that allow his clients to transform simple information into useful data that can be easily used.

In 2019, he co-founded the start-up LexRock AI. It is a “spin-off” of existing activities, talents, and technologies within the Irosoft company. It draws on more than 25 years of experience, expertise, and intellectual property, mainly in the legal world. LexRock AI is dedicated exclusively to the development and commercialization of artificial intelligence and language understanding products for various industries, such as the financial and insurance sectors, law firms and legal departments, and the cybersecurity community.

Alain has been very involved in the ICT ecosystem in Quebec for almost 20 years. He is currently involved with FORUM IA Québec (BOD, Treasurer), IVADO (stirring committee), MILA (startup committee), PINQ (BOD, Treasurer), the University of Montreal (ICT committee), FCCQ (BOD, president of the ICT committee), Groupe 3737 (BOD), Parc Olympique (BOD), and Centris (BOD). He is mainly known for defending the interests of SMEs in information technology while promoting the ICT field. He is also involved as a speaker and columnist (CSCience.ca) in order to raise awareness and democratize artificial intelligence. He is also very proud about LexRock AI being selected by Scale AI, the Canada’s AI Global Innovation Cluster, as part of a delegation of 12 Canadian companies and organizations to promote our national AI ecosystem at the World AI Cannes Festival (WAICF2023).

“We have built a platform where the human is always in the loop and where the artificial intelligence adjusts according to the human’s behavior.”

“We are recognized in Quebec as highly innovative and creative. This is what we have done to position ourselves where we are today with very modest means and an extraordinary team.”

“Our platform accelerates productivity, improves performance and organizational efficiency, while significantly reducing the cost of text and document processing. Some of our customers have saved up to 90% in time.”