The Company

Botminds Inc. provides intelligent document process automation for enterprises. Founded in 2015 by Gokul Ganapathi, the current CEO, and Ansari Ismail, Botminds has received funding from Creative Destruction Lab (CDL) in Canada. The company is headquartered in Seattle, Washington, has its R&D center in Chennai, India, and serves clients worldwide. Botminds built its own semantic engine on top of state-of-the-art algorithms and networks, and its stated goal is to change the way information-intensive, content-heavy, document-centric process automation is executed in enterprises.

Botminds says it is cash positive, and we expect it to move to a Series A round of funding in the coming year. The company has been actively selling its capture and automation products for the past few years, and these are the focus of this report.

The Technology

Botminds provides four critical components in its platform, offering:

• Data capture
• Search
• Analytics
• Automation

Together, they deliver an AI-first intelligent document processing (IDP) product set (see Figure 1). Though many IDP products are available in the market, Botminds is differentiated in two key ways. Firstly, it has built its own proprietary technology stack, including its own AI algorithms, which is unusual. And secondly, its focus is not on generic short-form invoices, statements, forms, etc., though it can certainly handle those. Instead, its focus is on long-form complex documents and web data sources with unstructured data, which, though not unique, is still relatively uncommon in the IDP world.

Digging into the Botminds platform, we find that this is a no-code system that leverages deep learning and natural language processing (NLP). Essentially, it’s an all-in-one, SaaS-based system for capturing or extracting, processing, searching, and analyzing data from documents or web pages and further automating workflows. But what particularly caught our attention is that sitting atop the core platform is a series of specialized and pre-defined solutions that can plug into the core platform. For example, there are plugins (or building blocks) for credit term agreements, KYC, LIBOR data extraction, etc. – in other words, mini business applications that can be deployed quickly and easily. It’s a Lego-style approach whereby you can either build your own solution or orchestrate and scale multiple solutions for more complex applications.

Digging deeper, we find Botminds is an AI platform that includes a wide range of AI and machine learning (ML) models. For example, one AI model can intelligently extract information from tables, another can extract insights from PDFs, and yet another provides semantic text understanding. So, it is a true platform for non-technical business users and SMEs as well as developers. At the same time, its out-of-the-box solutions/applications make it exceptionally flexible. Wrapped around all of these elements are customizable reports and dashboards for business users. One could, for example, easily configure a dashboard to provide weekly or real-time reporting. A lot of thought and design has gone into the use of the platform and all its moving parts, and as noted above, despite the underlying complexity this is all accessed and configured through no-code.

To fully understand Botminds, one needs to take a step back. Though arguably not unique, it is not in any sense a regular IDP system as those traditionally rely on a combination of scanners, optical character recognition (OCR), and bots (RPA). Instead, Botminds leverages its own AI tools to extract data from long-form, knowledge-rich data sources, be they web pages, PDFs, or tables. Important to note here is that no templates are required (as they are in traditional capture); understanding the zones, locations, and information on the page is fully automated by the AI within the platform. Once the data and text are extracted, the system runs an analysis to enable automation following specific rules and then to categorize, index, and sort the data so that it can be searched – again using its own cognitive search capabilities.

Botminds’ potential uses, then, are broad, ranging from financial services, real estate, and legal to healthcare. A user can build complex, proprietary, and powerful processing workflows that the company says should not take more than 2-4 weeks to be up and running. The examples given to us were drug safety research and background check verifications. Or, on the other hand, you could simply choose a pre-built option from the Botminds marketplace to meet your needs.