On-Demand Self-Serving Analytics
We have moved away from traditional data-warehousing concept of transforming and loading transactional data from OLTP (or other data sources) to OLAP, and querying the data-marts with proprietary languages like MDX. Instead, we just load (without transforming) the transactional data into columnar storage and create an interactive and complete visualization over the aggregation of the transactional data with the help of standard SQL queries. This approach has the following advantages:
- We are not dependent on specific technologies like MDX, instead we create interactive visualizations with the help of SQL queries, which gives more portability.
- As opposed to traditional data-warehousing tools, the design phase for analytics views which need customized design models (created by tool consultants) to create the views on. In Ideal Analytics, the users themselves (administrative as well as end users) can configure the data-sources and interactive views are created off-the-shelf.
- Another advantage of this approach is micro batching for frequent incremental update, which is practically impossible for traditional systems with OLAP storage for data warehousing.
Large Data Handling
Most of the existing BI tools in the market has the limitations on volume of data and analysing those in real-time. Ideal Analytics wins the race at ease in this aspect and can analyse millions of records in real-time with its unique data loading optimization technique.
Performance
Ideal Analytics provides lightning-fast performance with the help of:Ideal Analytics has the capability to load data to its column storage both on-demand (pull) and automatically by receiving the data (push).
- Columnar storage for analytics data for quick data retrieval
- Unique data loading optimization technique to minimize the data traffic
- In-built query optimization engine with intelligent caching mechanism
Data Load and View Update Strategy
-
Pull Strategy: This methodology is primarily used to reload the complete data-set. In case, the backing data-source allows installation of custom agents (e.g. a database agent installed in the backing RDBMS, which has the capability of providing incremental data-set), this strategy will cater for on-demand incremental data load.
- Push Strategy: In case, the transactional data-store allows for installation of real-time synchronizing agents, ideal Analytics can listen for updates in the datasets and update the data-sets on a real-time basis. The Ideal Analytics view has the capability of updating the view automatically, in case the underlying data-set is updated (as mentioned above). The view update frequency can be configured while configuring the dataset, according to the nature of the dataset.
Adaptability
Ideal Analytics provides an API suite for developing custom connectors to a proprietary data-source/system. As a result, analysis of data stored in a proprietary system becomes easy with quick turnaround time for development. It also provides out of the box implementations of industry standard platforms like Sales-Force.
Enterprise Scalability
Ideal Analytics servers are linearly scalable. Ideal Analytics instances are mostly stateless, backed by centralized and replicable instances of RDBMS storage and column storage and fronted by High availability proxy server and hardware load balancers. The only state-full nature of the server is the authentication state, i.e. login sessions, which are by default distributed cache enabled. So, a new Ideal Analytics server instance can be added to the existing cluster seamlessly, to scale up the performance.
Flexibility in Analysis
Ideal Analytics provides several layers of flexibility in analytics to suit diversified user requirement. Following are the some of the techniques it provides:
- Columnar storage for analytics data for quick data retrieval
- Unique data loading optimization technique to minimize the data traffic
- In-built query optimization engine with intelligent caching mechanism
- Data Load and View Update Strategy
Externalization
One-click externalization of dashboard items (for embedding in proprietary applications/web-sites/live documents, PDF, PPT etc.) makes it very unique and flexible. For example, a complete set of dashboard items can be embedded in a PDF document and the same can be distributed / shared. The document is automatically updated to reflect the latest data.
