Responsibilities & achievements

• Trained a commercial profiling machine learning model, to score loan applicants based on their attractiveness (repeat loans, returning). • Designed and implemented API to fetch features from customers about metrics like earnings, job type, credit score, financial products owned, to return a score. • Deployed the API creating the endpoints in Sagemaker and exposing them on API Gateway, to get an 37% increase Q4-20 vs Q1-21 on approved loans. • Analyzed patterns in selfies with the ID sent by customers to detect common fails like light overexposure, unreadable documents, face matching and fraud attempts • Trained a deep learning model on Keras to classify acceptable selfies, where the face of the holder and the face in the ID matched, and to avoid fraud attempts • Deployed the solution on EC2 to fetch the images from S3 and tag them acceptable or not, to reduce clerk manual workload in 82% in 2020. • Designed the data pipeline to fetch data from our databases in PostgreSQL, customer profiles from Mambu SaaS API and web usage from Google Analytics • Defined the performance metrics from commercial and financial perspective, customer behavior CSI scores and documented the metrics • Created the reports and dashboards using R markdown and flexhdashboard, designing visualizations and interactive graphics in JavaScript and Highcharter • Deployed the solution on EC2 using Shiny Server, on a virtual private cloud to be accessed by employees at Independencia, in time record (first month of Q1-2020)

Responsibilities & achievements

• Defined with the HR representative (Southern Peru Mine) the problematic of low engagement of technicians and the goal to improve attrition • Analyzed the data from technicians like education, employee satisfaction, income, promotions, work-life balance, and others to see what impacted employee turnover • Created a machine learning model to predict attrition for the next three and five years, with an accuracy of 83% • Defined with the HR representative (Southern Peru Mine) the problematic of low engagement of technicians and the goal to improve attrition • Analyzed the data from technicians like education, employee satisfaction, income, promotions, work-life balance, and others to see what impacted employee turnover • Created a machine learning model to predict attrition for the next three and five years, with an accuracy of 83% • Defined with the BI administrator (Divemotor Peru) the need of reports on parts sales, pricing, competitiveness, and opportunities to improve revenue • Analyzed time series of parts sales, market basket analysis, price comparisons between competitors and elasticity of demand • Created an automated report with prescriptive steps towards price of parts, where to apply promotions and which parts could rise the price to improve revenue

Responsibilities & achievements

• The challenge was to increase sales by 40% YoY for 2017, in a difficult environment of fewer vehicle sales and dealers in Peru • Asked for access to the SAP data, GM Maxis Platform, and Power BI Report Server to create a data pipeline and stream to an analysis database to a final report • Created a commercial dashboard mixing vehicle sales and historic parts sales on Power BI, with prescriptive steps for promotions in parts to improve revenue