Your browser is old and is not supported. Upgrade for better security.

Invest in KMEngine

KMEngine provides AI and NLP powered knowledge management for teams

Highlights

1
Knowledge management continues to become critical for companies as employees become more mobile
2
Helps companies recognize critical intellectual property and strategic knowledge, not just document knowledge
3
🔥 Founder has 251M in successful exits
4
Awarded Best of Show Activate 2019 (our core EnterpriseNLP product)
5
🌐 One Trillion Market Cap for Knowledge Management by 2027 and 20% growth each year (CAGR)
6
📈 Fortune 100 customers currently in production for over 200k employees
7
🧠Experienced team with senior roles at IBM, Oracle, Visa, prior CEO and CTO roles
8
Uses advanced NLP engine to find and relate data

Our Team

I come from a long history of applying AI and natural language processing to problems. We first developed state of the art search technology and believe that knowledge management provides a more complete higher value solution than just search. The field was open to improvement as other products in this space don't use AI and were outof date

The Story of KMEngine

We started in 2018.  The original vision was to do more intelligent AI and NLP language work and eventually build better conversational systems.  I was frustrated with Alexa at the time. It didn't follow continuous conversations I had to ask just simple one question one answer.  

Another genesis of the company was working on advanced models for knowledge representation using NLP and an architecture we call SULU - which is a more 3 dimensional graph/nlp hybrid.  Our early vision was to work on "digital brains" and become the intel of that space.  But we got a call to help Price Waterhouse Coopers with their search.  It wasn't working. They had all the top tools and AI and it still wasn't working. Could the top corporate search engines really be this bad I thought. 

So we developed a core search engine using NLP and what we call our Morphological Grammar Engine which uses morphological structures for comparisons, like comparing two snowflakes. It worked great and they saw increases in precision of 40-50%.  We took that core and worked on it for two more years and it was the birth of our ultra fast precise enterprise scale NLP engine - EnterpriseNLP.  We think it's the fastest most precise search engine on the planet (yes, even better than GPT-3).  And that technology became a core of KMEngine.  KMEngine is the simpler easier to use version of EnterpriseNLP combined with things specific to capturing experiential knowledge and some team integrations. It's also a SaaS not a on-prem install so it's much easier for companies to get started with, try before you buy, and the acquisition cost is much lower for companies to start out with. 

Ontologies that are truly Ontic

The second piece of the puzzle was building a tool to capture the language of the business. It was the primary reason why search was failing. We called it Ontology Builder and it listened to what users were searching for, ran it through an AI, and with a cooperative process with humans helped set up a language system. It increased search precision a lot.  Ontology Builder was first called "top search optimizer" because for the most common searches which are about 50% of search, it produced dramatic inceases in precision. It also solved the problem of routing internally facing and externally facing search questions to the right destinations, a major headache for companies.  

We won the LucidWorks Activate best of show contest in september 2019. It was a high point for us as we beat out a lot of the top tech companies that had many more millions to develop their technology.

But how do you prove you've got the best NLP search engine on the planet ?

We decided to do a live covid search on our website.  Only the best search companies had the courage to put out a live search and everyone was trying to help with the covid research effort.  We finished and we were ten times faster and had better answers for just about all the queries we would try.  For any of the tech companies BRAVE ENOUGH to put out a covid search we could compare them and always we were much stronger and ten times faster.  

A Critical Meeting with our First Advisor

We met Tony Rehm who is a knowledge management expert and immediatly loved what he had to say about taking KM further with AI. We showed him our search tech - "This is the future of KM" he replied. He was excited.  At the same time remote working and teams were just becoming the norm for everyone.  There was a good opportunity there.  So we plunged in learning everything we could about KM and all the other products in the space. Besides having terrible old fashioned search (some not even NLP! like they were still using Text Rank!) they just didn't seem to have it at all.  Building complex taxonomies like a librarian would - rather than  listening to the language being used.  And on the authoring side things were all wrong. People were treating knowledge management like writing an article (some products even called it an article). If you study the history of knowledge management like at Toyota you know they were working with hand written cards but focused on process and improvement by steps. So the first issue was the creation side of knowledge management was all wrong.  We reviewed it with Tony and he agreed. 

Shifting to Strategic Knowledge and Cognitive Applications

So typical KM companies like Verint or Lucidworks might supply a call center person with a piece of document knowledge.  And they might use some clunky degree of AI to do it.  Our EnterpriseNLP product already did that better and more precisely but we didn't want to play in the crowded customer support space.  

We decided to focus on capturing strategic knowledge. Strategies and rules of thumb that experienced workers would have, rather than a sentence from a document. This "experiential" knowledge is very hard for companies to capture today.  

Our background in AI helped us to focus on a cognitive application combining cognitive search but also an AI for relating information.  This could relate knowledge items ot existing knowledge items, patents, or information from a centralized EnterpriseNLP knowledgebase.  

Discovering Experts

Large companies often struggle with getting experts to talk across divisions or even how to recognize them. Some existing tools would process biographies or research papers. Instead we took a experiential approach by the knowledge items they submitted, how high they scored, etc.  We had a unique way to help build a database of experts cross organizations based on the ideas they had and the quality of ideas. I think that's a pretty strong proposition. 

Starting to think about TEAMS

So the last piece was the issue of knowledge getting stagnant.  Whats the point if it just sits there.  The current thinking is that you searched for your km.  A few were more dashboard like and you'd get notifications.  It wasn't really knowledge flowing around.  We were using SLACK at the time and saw they had an API to integrate. So I thought ... what if I can CREATE a knowlege piece in SLACK   and then later REFERENCE a piece in slack as well like a hyperlink? Now things were getting sexy.  Or you can have a specific filter update a SLACK channel with relevant knowledge postings.  

A Larger Search

One thing that becomes interesting is now when people would perform a cognitive search we could return not only our "potential answer" - single sentence result, but also include all of the strategic items related to that space - strategies from the most experienced people in the company related to that space. 

Honing the Pitch

We had to focus on the real pain point. And how the world was changing never to go back. We went from Mainframes to microcomputers to computers we call cell phones that are fully connected. Similarly work went from companies where you would work forever to places where the best minds were constantly hopping off for other opportunities. Working remotely just exacerbated things and we knew of several fortune 100 companies who were so disconnected that they were actually developing the same thing at three different places in the company. Strategy wasn't unified and from the bottom workers were task oriented but didn't know the bigger picture.  A small person might have the giant idea. How to get it to ping around and move through the company? Or from the top end, a company may wish to communicate a strategy down to all their teams. 

We decided to add our HIVEMIND technology so people could rate individual parts of the knowledge. Crowdsource it if you will.  We did thinking on what kinds of templates people would need - how knowledge should be structured. And so we created KMTemplates - the last piece of our puzzle. 

In the end KMEngine is the "System of Intelligence" that is the key value of the Corporation to execute.  Everything else is just engagement or storage. 

Recognizing Critical IP

One key aspect of our product is that everything that gets created is an IP candidate. Some might be taken and expounded, others patented. Companies have few systems to grow their IP and much of it simply leaves with the employees. We added the ability to flag items for IP potential. Later we might integrate IP search and show related items.  In the end, Knowledge Management is not a bunch of articles or a search engine or a taxonomy, not really any of the tools that the other KM companies are pushing. It's about recognizing critical IP and strategies.

Onwards from WeFunder!

This technology is revolutionary for Big Companies, Internationally distributed companies, Remote team distributed companies.  The workforce has changed and this tool can help tie the expanded world together. The Agile team management tools were all linear. This goal. These tasks.  We can be the most critical piece for a company and not just sit there but be an initiative technology to drive the company - any company - forward. 

We see ourselves as a core NLP technology company and there have been several large acquisitions in that space lately as companies realize this is a much less nebulous tech to get immediate results with than general ML. 

That's the story of KMEngine. We hope you'll join us on our journey. 


Downloads

Overview