Agustin Begue
Verified Expert in Engineering
Data Scientist and Machine Learning Developer
Buenos Aires, Argentina
Toptal member since October 4, 2021
Agustin is a versatile data scientist and machine learning specialist with a strong technical background in developing models and solving complex problems. He has a profound regard for the business impact he has and the value he adds. He also excels in solving challenging problems; adapting quickly to new environments, projects, and teams; and learning and applying new capabilities with ease.
Portfolio
Experience
- Data Analytics - 5 years
- Python - 4 years
- Machine Learning - 3 years
- Scikit-learn - 3 years
- Amazon Web Services (AWS) - 2 years
- TensorFlow - 2 years
- SQL - 2 years
- Django - 2 years
Availability
Preferred Environment
PyCharm, Python, TensorFlow, Scikit-learn, SQL, Amazon Web Services (AWS), Pandas
The most amazing...
...experience I've had was leading the data science team for an app designed to improve management and monitoring of physical assets using computer vision.
Work Experience
Recommender Model Data Scientist
Toptal
- Researched benchmark system architecture types and best practices for different recommender systems and designed the client's whole custom architecture for a video platform.
- Developed three models: (1) retrieval: given a user, scans the whole corpus of videos, getting a sample of 2,000 the user might watch; (2) ranking: sorts a sample of retrieved videos for a specific user; and (3) item-to-item: retrieves videos similar to a specific video.
- Developed a CI/CD pipeline for model training (with data preprocessing and hyperparameter tuning) and deployment. Data privacy needs required training pipeline to be fully automated as production data was not available to developers.
- Developed a CI/CD pipeline for model retraining (took deployed model and retrained specific layers) and deployment. Certain layers needed to be updated to receive newly subscribed users and newly uploaded videos.
Data Science Senior Analyst
Accenture
- Implemented four two-step computer vision models (object detection and anomaly classification) that detected 90% of anomalies and were deployed into production.
- Co-developed a pipeline for automated training and testing of computer vision models to enable fast assessment of overall performance, saving significant time for the rest of the team.
- Researched new business opportunities for Accenture and developed three business proposals to sell projects to existing clients and expand the value added by the company.
- Developed a demand forecasting solution for a large oil and gas company to adapt easily to different scenarios in the uncertain context of COVID-19. I was in the team in charge of predictions.
Data Science Analyst
Candoit Engineering & Technology
- Developed and deployed three tabular machine learning models for an insurance company, which helped them reduce churn, reduce costs, and reduce the number of trials they faced. All the models performed above 90% F1.
- Communicated with the client to understand the business, the problems they faced, and pain points. Proposed four machine learning-based solutions that would help them solve some of these issues, and two were implemented.
- Developed a forecasting solution to predict the number of students that would register for each class at a top university in Argentina.
Teaching Assistant for Introduction to Programming
Intituto Tecnologico de Buenos Aires
- Prepared and presented several classes thoughout the course.
- Tutored several groups of students on their semester projects, addressing doubts and issues they may have faced.
- Owned the process and corrected exams and assignments.
Experience
Bottle Vision
https://github.com/agusbegue/bottle-visionReal Estate Pricing
https://github.com/agusbegue/realestate-scraperEducation
Master's Degree in Industrial Engineering
The Buenos Aires Institute of Technology - Buenos Aires, Argentina
Certifications
AWS Certified Cloud Practitioner
AWS
Skills
Libraries/APIs
Scikit-learn, Pandas, TensorFlow, Keras, XGBoost
Tools
PyCharm, TensorFlow Serving, Amazon SageMaker, Jenkins, AWS CloudFormation
Languages
Python, SQL, Java
Frameworks
Django, Scrapy, Streamlit
Paradigms
Anomaly Detection
Platforms
Amazon Web Services (AWS), Docker, Amazon
Other
Machine Learning, Business Cases, Statistics, Data Analytics, Data Science, Cloud, Finance, IT Project Management, Web Scraping, Web Development, Object Detection, Computer Vision, Predictive Analytics, Image Processing, Web App Development, APIs, Recommendation Systems
How to Work with Toptal
Toptal matches you directly with global industry experts from our network in hours—not weeks or months.
Share your needs
Choose your talent
Start your risk-free talent trial
Top talent is in high demand.
Start hiring