Welcome to the first installment of the Truviso blog.    Truviso is in the business of Data Analytics.   It’s no secret that data analytics is a hot area in technology right now.   And it’s no secret why.

Businesses of all sizes are faced with unprecedented amounts of information.   As the world becomes increasingly interconnected through networks, every activity of every participant, whether a company, a department, or an individual user, generates streams of data that is crucial to their needs and goals; understanding and then quickly acting upon this information is the key to competitiveness and success across all industries.

The volumes of data that need to be analyzed are growing at a rate much faster than Moore’s Law and the other technology laws that in the past have allowed analytics systems to keep up.  But, as the pace of business and interaction increases, the pressure for more immediate visibility has increased.   These dual pressures – the growth of “big dataâ€� and the “need for speedâ€� – have caused the traditional approaches to data management to break down.   The result has been a tremendous acceleration of activity and innovation in Data Analytics.

Many emerging approaches are based on parallelizing the query engine.   Data processing has been long known to be a great use case for parallelism.   Teradata has been doing this for decades.   More recently, companies have been exploiting commodity hardware, and increasingly commodity software to lower the cost of parallel database systems.   At the same time, open source data parallel infrastructure such as Hadoop has been attracting a lot of interest as well.  The problem with relying solely on parallelism  is that it is really a “brute force� approach, and as such, with data volumes growing faster than hardware is improving – users of these systems are signing up for buying, managing, and powering an ever-growing complex of servers.  Furthermore, parallelism doesn’t help with providing answers quickly – in fact, it makes low latency processing harder.

Solving the Data Analytics problem requires more than just using more hardware to run more copies of the same old query engine.   Rather, what is required is a rethinking of how the query engine works in the first place.   This is where Truviso comes in.

Truviso’s Continuous Analytics is a different way of analyzing data, one that is tuned for the challenging analytics workloads of network-centric businesses.   As the name implies, Continuous Analytics is always on.   Rather than saving up data, then storing it to disk, then starting to process it,  Truviso takes advantage of the additive nature of data streams.   Queries run continuously and as new data arrives it is pushed directly through the queries.   Queries incrementally produce new results, which can then be immediately sent to dashboards, message buses or other applications, or can be stored in native tables for later access and reporting.

Continuous Analytics involves innovative stream processing technology, but is much more than that.   Truviso has a unique “stream-relational� architecture in which streams and persistent tables are unified to provide a seamless mechanism for querying the present as well as the past.   As a result, Truviso can solve the scalability problem for traditional reporting applications as well as enabling those applications to evolve into a more real-time mode of operation as business demands require and as business processes evolve.

That’s a quick overview of the underlying approach we are taking at Truviso.

Through this blog we will be describing our technology and its uses in more detail as well as looking at the evolving landscape of data analytics in general.     It is exciting to be in such a fast moving and crucial area and we look forward to sharing the journey with you.