| Architecture |
| Being a retail specific Business Intelligence solution‚ ARC architecture is designed to maximize performance while addressing retail specific needs such as data volumes‚ user bases‚ algorithms and time granularities. ARC performance oriented architecture is geared to address the advanced analytics requirement of retailers – addressing key requirements of performance‚ robustness and scalability. The ARC architecture incorporates the world's first Hybrid OLAP model (HOLAP) data model that integrates relational OLAP (ROLAP) and multi-dimensional OLAP (MOLAP) capabilities. |
|
 |
|
| Our architecture has been built to reflect a highly intuitive user interface‚ high performance and a single version of truth. Towards this the architecture comprises the following: |
|
| Model based Data warehouse and integrated Business Intelligence stack |
| ARC has an integrated architecture that realizes the power of a retail data model‚ retail algorithms and warehousing technology to drive relevant analytics. Its self service design brings in ease of maintenance through minimizing IT administration needs. ARC has retail-specific cubes and over 1000 reporting views pre-defined within the data model that can be accessed‚ re-configured or modified through its user interfaces. |
|
| As a result‚ ARC pre-configured solution easily connects to retail data sources‚ providing enterprise-wide analytics coverage from store to supply chain while drastically reducing deployment time. |
|
| Web-based front end |
| ARC web-based front end application that provides user interface mechanisms such as scheduled reporting‚ application specific caching of common data to avoid reloads from persistence storage‚ user specific caching which caches data for specific user sessions using the ARC portal cache‚ lazy loading technique or lazy initialization technique that enables on-demand loading‚ among others. |
|
| ARC achieves rich user experience through its easy-to-use‚ intuitive interfaces while maintaining a zero footprint front-end design. |
|
| Data warehouse design and aggregation strategy |
| ARC has a retail-specific Data warehouse design and aggregation strategy that brings down data warehouse size by 75% compared to general Data warehouse design by other Business Intelligence products. With pre-built KPIs and cube analysis‚ the aggregation strategy foregrounds data analyses and metrics that you use over a time period. This ‘Smart Aggregation’ technology is based on deep experience on how retail managers typically analyze data and wish to slice-dice data‚ drill down‚ across and through. |
|
| OLAP engine based on HOLAP technology |
| This revolutionary architecture that drives relational and multi-dimensional OLAP capabilities from a single platform‚ ensures that the item level details are maintained in the database with all aggregate information stored in its memory. Data stored in the memory is hence re-used‚ avoiding multiple executions of similar queries. The OLAP cache‚ disk cache and cold start are other features in place to ensure that there is optimized use of data accessed from the database‚ with strong dependence on data stored in the memory. |
|
| At a high level‚ this technology drives operational‚ tactical and strategic analysis from a single platform. |
|
| Database parameters |
| ARC table partitioning feature - PSA and non compressed F-FACT tables are partitioned by the system (through request ID) while the compressed E-FACT table can be partitioned by the user by certain time characteristics - effectively uses the partitioning functionality of various platforms. |
|
| For time range portioning‚ ARC manages partition functionality in its core architecture as well. ARC DB optimizer decided the access path based on information provided by database statistics. Pre-defined database parameters for given hardware configuration and data volume anticipation are in place. ARC has been well-tested on various platforms with varying data volumes. |
|
| ARC hence brings you a highly scalable environment both in terms of functional analytics and data volumes. |
|
| Talk to us for a more detailed discussion on ARC architecture. We would be glad to share technical information. |
| |
