Keep in touch with meI'm using Intch to connect with new people. Use this link to open chat with me via Intch app
Work Background
CompassMachine Learning Scientist
Oct. 2023
TryolabsMachine Learning Engineer
Mar. 2022 - Oct. 2023Montevideo, UruguayMachine Learning Engineer at Tryolabs, working mainly in the areas of forecasting and pricing. Worked as one of the main engineers designing, developing, and deploying a dynamic pricing system from scratch for a New York-based used electronics company. The project achieved a statistically significant uplift in gross profit as measured by the causal impact method during the first 6 months of work.
ScanntechData Scientist / Team Lead
May. 2020 - Feb. 2022Montevideo, UruguayTeam lead within the Data Science branch at Scanntech, a rapidly expanding Uruguayan based company that's leading tech innovation in the retail industry in Uruguay, Brasil, and growingly in the rest of LATAM. In the Brazilian market Scanntech handles the data of 70% of the transactions that take place at retail level; billions of rows a month. As team lead, I was in charge of setting up a product classification process that transformed this raw data, usualy in text form, into databases that were critical for the quality of all the further products that were sold to the clients in one way or another. I was in charge of setting up the process from end to end, which involved: * Data extraction using SQL and Spark. * Feature extraction using NLP tools. * Classification model development and finetuning using sklearn, tensorflow and pytorch. * Model availability using FASTAPI. * Batch Processing and orchestration using Airflow. * Team management using agile methodologies.
EntrepreneurEntrepreneur - Data Scientist
Jun. 2019 - Dec. 2021Montevideo, UruguayDeveloped a crime hotspot forecasting system for the Montevideo police department. More precisely, the goal of the project was, after dividing the city into grids, to forecast the likelihood that any given grid on any given day would become a peak for criminal activity, i.e. a crime hotspot. This type of problem was usually tackled in the literature with stochastic processes, more specifically self-exciting point processes. For this project, I used a niche deep learning architecture called conv-lstm, which embedded the convolution operations inside the lstm cell, and was particularly well suited for spatiotemporal modeling. The POV managed a 50% recall for 25% precision.
LUNAROM S.A - MiEdificioCo Founder
Apr. 2017 - Mar. 2019Montevideo, UruguayCofounder of the SAS platform MiEdificio, oriented to the real estate market. The platform was designed and developed from scratch and launched to the market, reaching over a thousand users. The project received support from the business incubator Ingenio and financing from the ANII agency. The product was ultimately developed to the satisfaction of the financing organization and of its first clients but was eventually abandoned due to reevaluation of market size and opportunity for the product.
Intch is a Professional Networking App for the Future of WorkSee more people like Guzman on Intch
300k+ people
130+ countries
AI matching
Get Intch if you'd like to
453430 people
15
Consulting Practice Manager
15
--------- @ ------------------
17
CEO @ LogiKom
35854 people
20
Analista de sistemas @ ANTEL
20
Data Analyst @ PROSEGUR
17
Data Scientist @ Interflora