Data Management

Uniquely, Mercato has automated the process of sourcing and cleansing raw 'dirty' data from any dimension, including Lotus notes, through a process of:

• Fetching and receiving data feeds - multiple formats, languages, proactive and reactive sourcing, flexible rules to model business process logic.
• Filtering them through a data conversion module that converts myriad formats into a 'clean' matched standard using a sophisticated set of matching tools & algorithms designed to engineer the highest match rates possible
• Checking product data, assigning product codes and filtering new products for preparation.
• Comparing product data on a like-for-like basis eg. price, stock.

This is the basis of our Catalogue Management, structuring and populating having verified descriptions and formatted to a uniform style.

Once a database has been centrally developed, we apply Enterprise Transformation Loading to cleanse, transform and filter the data. This is the process of generating data from data, creating new data before placing into a Data Warehousing environment. These provide the repositories form which Business Intelligence can be collected and populated back to end users.

In combination with our Data Warehousing methodology, we apply a framework and architecture that offers ubiquitous delivery and a set of acute business rules to drive transformation of the data. This drives complete operational excellence and data integrity.

When the data underpins a tool from the Mercato Application Suite, we apply heuristics to automatically measure end user activity to drive better data quality through rules.

Complex data structures and algorithms are used to manage and unlock huge data 'search' level potential to an end user. Where appropriate we deliver parametrics on data. This is the forward automated process or flow of taking meta-data and allying it to consistent related attributes and ranges. This unlocks intuitive referencing and searching capability - easier, quicker, more accurate and guided.

The successful enterprise effectively extracts only the data required to make a decision and quickly turns that data into action. However, there are a number of key challenges along the transformation round map:

• Sourcing data feeds, delivering data models, enforcing quality assurance, mapping data, building data and cleaning data
• Cost of bad data quality to the organisation
• Managing huge volumes of regularly changing data
• Validity and security
• Deciphering between and managing appropriately the two core types of data:
  • Inactive data does not change over time, it is primarily the architecture that provides the context in which dynamic data is populated eg. Product specifications or photographs
  • Active or dynamic data is constantly changing and is usually a numeric eg. Price, period, stock, volumes
• Source data is incomplete, often misclassified, usually dispersed across multiple, incompatible systems and not standardised. Transformation demands detailed, accurate, comprehensive data
• Instilling a logical view of data and re-usability