Financial-services companies are under enormous pressure to process ever-larger volumes of transactions in near-real time, and their IT infrastructures may not be up to the task. But a next-generation database with data store, data analysis, metadata management, and transaction-processing capabilities could provide a solution.

Aleri Inc., in New York and London, has developed what it calls vector-based data-management software that can process millions of transactions in a fraction of the time today's database systems require. The software is in production at the Allied Irish Bank in Dublin and is being implemented at several other financial-service companies, says Aleri CEO Alan Hambrook. Utility and telecommunications companies also are potential users, as are any businesses with huge amounts of data, high transaction volumes, complex business logic, and sophisticated data-analysis applications, Hambrook says.

The software processes in a few minutes transactions that can take a legacy system 14 hours to run in batch mode, he says. It employs vector-processing technology, which uses built-in instructions to perform multiple calculations simultaneously, and memory-mapping file technology, which links the database to memory in the hardware. Business logic and algorithms that process data faster than those in relational databases also boost Aleri's software speed, Hambrook says. The database is further enhanced

by an embedded transaction-processing engine and metadata management capabilities for data analysis. The system even uses, on average, 30% less disk-storage space than other databases, because it doesn't have indexing tables.

For the immediate future, Hambrook envisions Aleri's software co-residing with relational databases, acting as a high-performance transaction processing and data-analysis engine and pulling information from otherwise fragmented IT systems to help companies, say, get a complete view of their customers.