William Suzuki

Software Engineer

LG Electronics•  February 2022 - Present

Implementing the IEEE Standard 1609.2.1, which specifies a Vehicular Public Key Infrastructure (VPKI). In the back-end I use Python with Flask-RESTful. I also need to implement and call C programs from Python using ctype, to deal with ASN (Abstract Syntax Notation) encoding. Knowledge of cryptography: AES, Elliptic Curves, Public Key cryptography, hashing. And use of pytest and makefiles to keep the code always in a working state.

Data Scientist

LG Electronics•  October 2021 - January 2022

Responsible for research and implementation of misbehavior detection models and anomaly function generation for vehicular sensor and position data. Tech-stack: Python, Pandas, pymongo, mongosh, plotly, bash, tmux. One of the main authors of the paper "Anomaly Modeling for Misbehavior Detection from V2V Basic Safety Messages", submitted to IEEE Transactions in Vehicular Technology. This research relies on domain knowledge in V2X, kinematic laws applied to vehicles, anomaly detection and machine learning.

Data Engineer

Tabas Real Estate Technology•  May 2021 - September 2021

Responsible for building web scrapers of major Brazilian Real Estate websites, I used selenium and python. Maintain routine crontab jobs, for internal data transformation and web scrapers. Bash, crontab, vim, tmux. Built Real Estate rental pricing models. I implemented an adapted KNN, GAM, xgboost and random forests. ch-stack: bash, crontab, tmux, vim, bash scripting, pandas, gspread-pandas, pytest, sklearn, pytest, streamlit. Responsible for implementing Airflow in GCP and migrating jobs from crontab.

Data Scientist

Tokio Marine Insurance•  February 2020 - April 2021

Implemented a clustering method using Python, k-means, Gaussian Mixture Models for grouping ZIP codes, the groupings would be used as categorical variables for the insurance price model (GLM). Visualizations were made using Bokeh. Automation of weekly reports that was done manually involving outlook and excel. I used pywin32, openpyxl and pandas for excel manipulation. Data preparation, feature engineering for pricing and probability of client churn. I used Vaex, MongoDB, pymongo, pyodbc, SQLServer, psycopg2, AWS Redshift. Advocated and implemented a POC of Airflow at the department.