2009 top underreported technology stories:10. Enterprise wikis become a platform
When engineers at Cisco Systems realized the networking giant needed a better system to track network errors, it turned to an unexpected technology: the wiki. What they built was no simple, collaborative Web page. It's an inventory-tracking application that correlates network problems with changes to network hardware and configurations, built on technology from Twiki, an open source software developer.
The wiki has long since become an enterprise tool. But now, wikis like the ones at Cisco are moving beyond their traditional role and becoming a platform in their own right. The applications are light, quick, easy to deploy, and far cheaper than traditional enterprise software. Custom-tailored wikis allow developers to automate workflows, pull data from multiple applications, and blend structured and unstructured data in innovative ways. What's more, wikis are beginning to include enterprise-type features such as permissions, versioning, and tracking, which makes it much easier for old-line IT managers to accept.
Consider this use of the new wiki: One major enterprise (the company asked that its name not be used) uses wiki technology from MindTouch to pull together a rich mashup of Sugar CRM data, trouble ticketing, and accounts payable. The information lets a sales rep or technician know right away if a customer pays on time or has a history of related network problems. Should a sales rep want to develop a new relationship, she can find out right away who at the company is connected to her via LinkedIn, for example. The system was developed fairly quickly using some of the 100 or so application extensions built into the MindTouch platform.
While we tend to think of wikis as collaborative documents generated by data entered on a Web page, Cisco's Twiki-based inventory system is integrated with SNMP, which allows IT to see changes to the hardware configurations in real time. A separate Twiki app at Cisco uses standard connectors to pull information from multiple databases -- or if the information only resides on the Web page, it will scrape the data from the page.
That wouldn't work if developers at Twiki hadn't taken a revision control system typically used to manage code and turned it into a database-like feature that can manage unstructured information.
Even so, extracting the full value of unstructured data is still quite a ways off, says Twiki CEO Jitendra Kavatheka. "Unstructured data represents human thought, and software hasn't caught up."