Sales Data Analysis
Budget: $30 – $250 CAD
We're looking for a skilled data engineer/data analyst to help with a comprehensive data processing and analysis task using a large liquor sales dataset (2012–2020). Below is a breakdown of the work involved:
1. Data Ingestion & Preparation
Import the "Liquor Sales" dataset into AWS RDS, setting up the schema based on a provided data dictionary.
If needed, split large files into manageable chunks to optimize performance.
Use Apache Sqoop to migrate the data from RDS to HBase, designing a suitable schema with appropriate column families and row keys.
Ensure data consistency and validate the integrity of the data in both RDS and HBase.
2. Data Cleaning
Clean the dataset to improve quality and ensure accurate analysis:
Remove incomplete or missing data.
Fix formatting issues and invalid entries.
Standardize categorical data and normalize field formats.
Deduplicate records to maintain clean datasets.
3. Batch Processing with MapReduce
We're looking to extract actionable insights through batch processing:
Revenue & Sales Analysis:
Calculate total revenue per store.
Identify top-selling liquor categories based on bottles sold and sales in dollars.
Aggregate and analyze sales performance at the county level.
Store & Vendor Performance:
Rank stores based on revenue, sales volume, and average transaction value.
Evaluate vendor performance by revenue and volume using vendor-related fields.
Time-Based & Trend Analysis:
Analyze monthly and yearly trends using the date field to identify seasonal patterns and long-term growth.
4. Insights & Recommendations
Provide practical, data-driven recommendations such as:
Optimizing store operations and sales strategies based on performance.
Identifying and promoting high-performing liquor categories.
Enhancing vendor management by recognizing top vendors.
Tailoring marketing efforts based on time-series and geographic sales insights.
1. Data Ingestion & Preparation
Import the "Liquor Sales" dataset into AWS RDS, setting up the schema based on a provided data dictionary.
If needed, split large files into manageable chunks to optimize performance.
Use Apache Sqoop to migrate the data from RDS to HBase, designing a suitable schema with appropriate column families and row keys.
Ensure data consistency and validate the integrity of the data in both RDS and HBase.
2. Data Cleaning
Clean the dataset to improve quality and ensure accurate analysis:
Remove incomplete or missing data.
Fix formatting issues and invalid entries.
Standardize categorical data and normalize field formats.
Deduplicate records to maintain clean datasets.
3. Batch Processing with MapReduce
We're looking to extract actionable insights through batch processing:
Revenue & Sales Analysis:
Calculate total revenue per store.
Identify top-selling liquor categories based on bottles sold and sales in dollars.
Aggregate and analyze sales performance at the county level.
Store & Vendor Performance:
Rank stores based on revenue, sales volume, and average transaction value.
Evaluate vendor performance by revenue and volume using vendor-related fields.
Time-Based & Trend Analysis:
Analyze monthly and yearly trends using the date field to identify seasonal patterns and long-term growth.
4. Insights & Recommendations
Provide practical, data-driven recommendations such as:
Optimizing store operations and sales strategies based on performance.
Identifying and promoting high-performing liquor categories.
Enhancing vendor management by recognizing top vendors.
Tailoring marketing efforts based on time-series and geographic sales insights.