These days, if people have a question they turn to Google for an answer. Startup Paglo wants to become Google for IT administrators.

The company, which launched today, is developing a search engine that lets IT professionals query their own network infrastructure and software.

Interested users can sign up for a private beta, which is expected to launch at the end of November.

Here's how it works. Customers download Paglo's crawler to a server or PC. The crawler gathers information about each machine on the network, including device type, name, IP address, installed software, and so on. The crawler communicates the information to Paglo's data center, where it's indexed and stored.

Customers then access their accounts to run searches on the information gathered by the crawler. In the demo that CEO Brian de Haaff showed me, he imagine an admin that wanted to know if he had enough Office 2003 licenses for all the copies of the software running in the organization. He queried the search engine for all the instances of Office on the network. The search engine kicked back the answer, which could be displayed as text, or in charts or tables.

Of course, any half-baked network software tool should be able cough up basic information like this. De Haaff says the value of the search engine is that it pulls this information out of its respective tool-based silos and makes it available with a simple query.

However, the tool has limitations. In the demo, de Haaff couldn't ask "How many licenses for Office 2003 do I have?" because the system can't yet discover that information. Instead, he assumed the administrator would know how many licenses he had, and then match it to the number of instances of Office actually on the network.

It would be great to see the crawler and indexer develop enough sophistication to query data stores, such as those held by an asset management tool, and include that information in its search.

Also, the results of a search will be influenced by the initial search request. A poorly worded query will generate poor results. Paglo provides two search options: standard keyword search, and the Paglo Query Language. The query language is intended for sophisticated searches, but also imposes a learning curve on would-be searchers.

Please note that the company itself is only five months old, so these critiques should be taken in context.

DID I MENTION IT'S FREE?
Paglo takes a page from the Web 2.0 playbook by offering the crawler and search service for free. "When we have a lot of users, then we'll think about monetization," says de Haaff. I'm sure his investors have given it some thought, especially considering that Paglo is absorbing all the costs for storing and processing each user's network data.

Advertising or subscription fees are the most likely business models. But de Haaff says that for the foreseeable future, Paglo won't charge for the service.

That's not the only Web 2.0 play. De Haaff also hopes to build a community of users that will share useful queries with one another and add features to the crawler (which is open source).

For those worried about data privacy (and that should be everybody), Paglo partitions each customer's data to prevent the information from being seen by others.

Pedigree is always a consideration with startups. Paglo's got a solid one. Co-founders de Haaff and Dr. Chris Waters come from Network Chemistry, a network management and wireless security vendor. They recently sold Network Chemistry's wireless security business to Aruba. They are using the proceeds, which are undisclosed, to finance Paglo. Paglo was formerly known as Project Wishbone.