Use Web Scraping to fetch information from a bookstore catalogue.
The Coffee&Cookies Books Club is a fictional startup in the UK that wants to build a subscription-based book club for lovers of coffee and books. They pick only hard-to-find and greatly-rated books. The company packs them with a unique reading guide and a wellness set including exotic varieties of coffee and specialty candies.
Coffee&Cookies needed to decide which books to deliver next to their customers. For that, they needed to understand what regular bookstores were offering.
I developed a datascraping model using python, jupyter notebook, beautifulsoup and requests to retrieve Books Titles, Prices, Ratings and Stock Availability.
Discover data and build views over a legacy “black-box” Product Lifecycle system using Snowflake and Tableau.
This company invested over € 1.2M in a legacy software to manage innovation and needed specific reports over the data in the system, but the provider would not develop them nor handle the data schemas.
The mission was to understand the data structure (data-mining) and ETL the relevant tables to Amazon S3 and Snowflake to build the views on Tableau, along with the data squads.
The relevant data was migrated to Amazon S3 and made available on Snowflake. We were therefore able to build the needed views on Tableau and set up a self-service BI for this platform. Saving now over €1.500 per report created.
Find opportunities to reduce the returns of sales with descriptive analysis.
This company recently expanded its operations to a new DC and within 6 months their inventory for Returns was reaching 80% capacity and growing.
I expanded the outlet store operations, offered shipping services from the outlet store, partnered with B2B buyers, and created an improvement committee to reduce the intake of returns due to transportation issues and customer dissatisfaction.
Inventory occupancy kept steady, the outlet store sales grew by 63% (compared to 2019), reduced markdowns from 75% to 67% (average). In three months, reduced returned items due to picking errors in 71%.
Other Operations Projects
Use Radio-Frequency (RFID) tags to improve footwear distribution in the United States
This company had problems with the availability of products in third-part point of sales in the US. There was a team of sales representatives the visited the stores once a month to put purchase orders, but it wasn’t enough and sales were being lost.
My task was to develop a viable way to remotely check the sortments, on a daily basis, to improve the operations intelligence and increase the product availability in the points of sale.
The pay-off was that we didn’t need sales representatives anymore, only replenishers, so the 10% commission (about USD 1.5/pc) was replaced by a cost of USD 0.08/pc (tag+system cost).
Find the root causes of production losses to increase yield in a footwear factory
This company saw its scrap and reprocess rates (losses) triplicate in a matter of weeks after new raw materials were introduced in the production system.
My mission was to urgently find which of the 23 ingredients or 10 processes were causing the production yield to vary, and find a way to minimize the problem.
In a few weeks, we set a costless MVP for a traceability scheme using Google Suite (Forms, DataStudio) to keep track of the raw materials being sent to production.
With the findings, we reduced the reprocess rate from 10,7% to 6,6% (38,6% reduction) in two months, with savings over € 2.1M/month.
