Personal details
Ricardo V.
Summary
I am Bachelor and Master Candidate at Universidade de São Paulo (USP), based in Brazil, with 5+ years of experience in technology careers with data driven projects and 3+ years of work as a Business Intelligence and Data Science Analyst, participating as Data Engineer, Data Analyst and Tech Lead in several projects, recently more focused in the Public Sector.
Currently, I am enrolled in a Master Degree research focused in football analytics, with focus on maximization of expected goals through match events of a team.
The main goal of my career is to allow technology to promote evolutions in other knowledge areas.
I truly believe mentoring is a way to spread my knowledge and enhance opportunities for programmers who might be stuck in several questions: coding, networking, portfolio and carrer development.
Work Experience
● Activities on demands of CNJ (Brazilian National Council of Justice);
- Construction of ETL to attend 27 courts to find inconsistencies of 3 lawsuits data sources, creating automated presentations, dashboards and machine learning model to infer the main themes of these lawsuits;
● Creation of qualitative-quantitative research over AI usage with results evaluation applied on a sample of 94 brazilian courts;
● Development of LLM and NLP models in order to reduce work time from almost 10 times less in tasks applied in lawsuits;
● Follow-up and deploy of clustering machine learning models from 3 projects made by partnerships between courts and universities;
● Management of Merchandising business teams demands, performing as a tech lead;
● Development of diagnostic analysis of most important sales days, reducing from 6 to 3 the number of reports;
● Collaboration in a new machine learning model for key accounts composition, aiming their company monthly sales rising from 45 to 52%;
● Processes automation, decreasing from 61 to 31% of working hours over time spent on routines and frequent reports;
● Development of new demand forecasting project applying Multiple Non-Linear Regression decreasing prediction error at product level from 100% at average, to a range of 10-20%;
- Business Intelligence solutions with clear analysis on mid-tail and incubation sellers data;
- Creation of queries and automated reports to track sales performance, selecting 20% of sellers with the best performances to become potential key accounts;
- Development of fuzzy matching algorithm to promote mid-tail sellers products, according to top sales products in the ecommerce, suggesting 10% of promotion prices