Search Boundaries. Explode.

November 14, 2013

I read a quite remarkable news release. The title? Grab your blood pressure medicine because you may “explode.”

Expertmaker: Artificial Intelligence (AI) Explodes the Boundaries of Enterprise Search

I expect a sign to warn me off. Was it safe to read about such a potentially powerful technology?

Expertmaker Info

Straightaway I poked through my information about search vendors. I did not recall the name “Expertmaker.” I think it is catchy, echoing the Italian outfit Expert System.

Expertmaker is located at  The company offers the following products:

  1. Consulting
  2. Products that are “an online solution and/or mobile solution.”
  3. Big Data Anti Churn. I am not exactly sure what this means, and I did not want to contact Expertmaker to learn more.
  4. Flow, a virtual assistant platform.

The technology is positioned as “artificial intelligence.” The description of the company’s technology is located at this link. I scanned the information on the Expertmaker Web site. I noted some points that struck me as interesting, particularly in relation to the news release that triggered my interest. (Who says news releases are irrelevant? Expertmaker has my attention. I suppose that is a good thing, but there are other possible viewpoints too. My attention can be annoying, but, hey, this is a free blog about going “beyond search.”)

First, the label “artificial intelligence” is visible in the description. The AI angle is “machine learning and evolutionary computing.” The point is that the system performs functions that would be difficult using an old fashioned database like DB2, Oracle, or SQL Server. (I assume that the owners of these traditional databases will have some counter arguments to offer.)

Second, the system makes it possible to build search-based applications. (Dassault Exalead has been beating this tom tom for six or seven years. I presume that the Cloud 360 technology is relegated to the user car lot because Expertmaker has rolled into the search dealership.)

Third, a development environment is available, including a “Desktop Artificial Intelligence Toolkit.” There are “solvers.” There are various AI technologies. There is knowledge discovery. There is a “published solution.” And there is this component:

Semantic, value based, meta-data structures allow high precision understanding and value based searches.  With the solution you can create your own semantic structures for handling complex solutions.

Okay, this is pretty standard fare for search start ups. I am not sure what the system does, but I looked at examples, including screenshots.

Here’s one:


One of the entities mentioned on the Expertmaker is Wiiable. The url for resolves to this page:


The News Release and Explodes

The news release is really about a profile created by one of the azure chip consultancies. The subject of the report is ostensibly “about” the importance of artificial intelligence in “enterprise search.”

The methodology used to validate “explode” is called Stratecast’s Data Management Model. I worked at a couple of high profile consulting firms as a full time employee and have worked as an advisor for many years. I have to admit, I am have not heard too much about the Stratecast Data Management Model in the context of enterprise search or artificial intelligence.

But the academic puff pastry is wrapped around the ground beef of the news release. The subject is that search has a problem. (I will reveal the solution in a moment.) The news release points out:

What is wrong with search is what will be wrong with it soon, as users who demanded, and received, sub-second search results turn their sights toward a more customized search experience—one that makes them truly feel as if they are no longer obtaining search results for the masses. Users are starting to demand increased access to the kind of expert and specialized knowledge that currently resides inside corporations and in online forums. Needed, instead, will be specialized search services capable of being tailored to an individual user’s preferences and behaviors. Users are coming to expect some analysis, or at least responses, which address what the user means, not what they said (or typed into the search engine). So, users may now expect search to be more of a query; e.g., “Tell me what I need to know about XYZ.”

Yep, I had to reread the passage as well. I think it means that old style search systems are useless. Specific solutions to quite specific information retrieval problems are required. The fix is for software to know what the human “means, not what they (sic) said.” Yep, software can do what words cannot. No problem.

The ground beef within the report is—hang on, folks—Expertmaker. What I find interesting is that I can link to an outfit that will sell me this document. I assume I will not explode if I read it.

Net Net

My view of this type of marketing is mostly neutral. Perhaps a better word is indifferent? I wrote a short item that points out that Hewlett Packard is marketing Autonomy IDOL in some interesting ways. I just received an email about “intelligent content.” I endured talks at an enterprise search conference that underscored the gap between the problem of finding information to wackiness.

Check out the report and the “beef” yourself. I am going to go back to my hobby of posting profiles of search and content processing vendors that make interesting case studies. You can find these at I think there are some marketing lessons in most of the write ups. I did not use a Fancy Dan model, just plodding information collection and common sense. I await the explosion as well.

Stephen E Arnold, November 14, 2013


Comments are closed.

  • Archives

  • Recent Posts

  • Meta