CASE STUDY

AUTOMATED DATA PIPELINE & BUSINESS INTELLIGENCE ARCHITECTURE

CLIENT PAIN POINTS

The Challenge

One of the world’s largest wine and spirits sellers was facing challenges with their data management process, from data acquisition through to business analysis. Most of the process was done manually by their sales team, leading to heavy productivity loss, transformation errors, data inconsistencies, and a lack of granularity preventing them from fully understanding their business activities.

EFFORTS

Our Solutions

  • We conducted an audit of their data mgmt. (including data acquisition from 15+ POS systems, transformation process done via excel macros etc.)
  • We provided a set of recommendations including how to automate the data acquisition process (based on the POS functionalities) and how to transform the data via code.
  • We designed a simplified and business-driven database architecture (data schema), keeping data granularity at the outlet & day level.
  • We built 4 proof of concepts (PoCs) based on recommendations, using RPA tools (selenium, UIPath), and integrating with their Azure Datalake, Databricks Guardian, and Snowflake.
  • We later helped them implement these solutions to production across all 15 retailers.
  • We worked with their local and regional teams in order to ensure proper knowledge transfer and have the client able to manage the new architecture independently.
  • We built 9 business intelligence dashboards, using powerBI, and fully integrated with snowflake.

OUTCOMES

The Results

  • The client was able to reduce the involvement of their sales team from a couple of days per month to just a few hours.
  • Thanks to the new architecture, they could obtain more granularity with sales data at the product-day-location level where it was previously aggregated
  • The number of errors dropped, requiring less maintenance, faster access to the data, and offering a more sustainable solution more robust to POS changes.

OTHER CASE STUDIES