Startup Of The Week: Attivio Offers A New Approach To Enterprise Search

Your company collects terabytes of data, yet finding the information that employees need is tricky because it's scattered far and wide. IT pros have tried many techniques to solve this data management dilemma, including universal databases, data warehouses, and enterprise search. Newcomer Attivio is about to offer another alternative.
-- John Foley ATTIVIO HEADQUARTERS: Newton, Mass.

PRODUCTS:Active Intelligence Engine, an enterprise search engine that combs through structured and unstructured data

PRINCIPALS: Ali Riaz, co-founder and CEO; Sid Probstein, co-founder and CTO; Andrew McKay, co-founder and senior VP of products

INVESTORS: $6.2 million in private equity funding

CUSTOMERS: Boldfacers

SEARCHING FOR ANSWERS

Businesses generate both structured data, that within the fields of databases and business applications, and unstructured data, the free-flowing content of e-mails and documents. However, database query tools and search algorithms generally don't work across all data sources, so getting comprehensive, timely information can be nearly impossible. THE MARKET Enterprise search is a fast-changing market in which Google, IBM, and Oracle compete with specialists like Autonomy, Endeca, and Vivisimo. In January, Microsoft announced plans to acquire Fast Search & Transfer, an enterprise search vendor in Norway, for $1.2 billion. Attivio's co-founders all worked at Fast, Riaz as COO and president. Does the world really need another enterprise search company? Try to find information you need inside your company, and you'll know the answer. WHAT'S NEXT Attivio is fine-tuning its software and signing its first customers. Early adopters are evaluating the product, with general availability planned for summer. Monthly license fees are expected to begin at $5,000. HYBRID APPROACH Attivio's Active Intelligence Engine is enterprise search software with characteristics of a business intelligence query tool. It works by creating a "universal index" of data from relational databases, Microsoft Office documents, PDFs, e-mail, and enterprise applications, including Salesforce.com. Among its features are a "fuzzy" query language, probabilistic relevancy model, real-time processing, and an alert system that notifies users when new results match a search request. There's an API for custom development. TIMELINE