ema = DataScientist( name="Sergio Emanuel Tomas", role="Sr Data Scientist @ AB InBev", stack=["Python", "PySpark", "Databricks", "SQL"], passions=["AI", "hard problems"],) # what I do every dayema.solve(hard_problem)Out[1]:
Sergio Emanuel Tomas
Senior Data Scientist at AB InBev
I solve hard problems with data, machine learning and AI, and take the solutions to production.
Paraná, Entre Ríos, Argentina. Open to relocation.
About me
I’m a Senior Data Scientist, and what I enjoy most is solving hard problems: the ones without an obvious answer, where you have to understand the business before writing the first line of code.
I love technology. Every day I research and try new things in tech, and right now I’m fascinated by how fast artificial intelligence is growing. I want to keep learning and growing along that path.
Before working with data I was a software developer, so I’m comfortable across the whole journey: from analysis and modeling to the pipeline that runs it in production.
- In tech
- 6+ years
- In data
- 5+ years
- Based in
- Paraná, Entre Ríos, Argentina
- Open to
- Relocation and international roles
Experience
present
Senior Data Scientist
AB InBev
- Built two internal chatbots: one guides new employees through onboarding, the other helps people understand a new in-house development platform.
- Lead the team technically on offline-channel initiatives focused on salespeople and their behavior.
- Trained and onboarded several new members of the Data Science team.
- Act as a technical reference for the data team across the company.
- Python
- PySpark
- Databricks
- SQL
- Delta Lake
- Airflow
Data Scientist
Datalytics for AB InBev
- Led and built an algorithm that optimizes customer visit times through four modules: customer relevance, competitive opportunities, service level and customer independence.
- Led, built and automated an algorithm that prioritizes and sends commercial actions to each customer based on purchase propensity, commercial KPIs and history.
- Detected fraud in one of the company’s most important promotional campaigns, generating significant Trade Spend savings (2022).
- Improved the allocation of commercial assets by measuring the benefit of placing them in different sales channels (2023).
- Contributed to a campaign-adherence probability model that helped increase customer participation (2022).
- Python
- PySpark
- Databricks
- Machine Learning
- SQL
Data Engineer
Datalytics for AB InBev
- Built, validated and documented ETL processes, dimensional models for Power BI and tables for the Data Science teams.
- Rebuilt trust with the team’s Business Partners, who were initially skeptical of the data area.
- Drove agile ways of working in the two teams I was part of.
- SQL
- Spark
- Azure Data Factory
- Azure Data Lake
- Power BI
- Presto
Software Developer (Trainee to Mid-level)
GlobalLogic formerly Hexacta
- Built APIs and frontend components, with tests and documentation.
- Delivered significant performance improvements to the application.
- Ran an internal training on RxJS and observables in Angular.
- Built SSIS ETLs and data migrations on production tables.
- ASP.NET Core
- Angular
- TypeScript
- Vue
- SQL Server
- SSIS
Featured projects
These projects were built at work. For confidentiality I describe the problem and the outcome, not the code or the data.
Internal AI chatbots
Built
Two conversational assistants for the company. One guides new hires through onboarding; the other helps people understand a new in-house development platform without needing someone to walk them through it.
Trade Spend fraud detection
Built
Detected fraud in one of the company’s most important promotional campaigns. The finding generated significant savings in trade spend.
Commercial action prioritization
Led and built
An automated algorithm that decides which commercial action each customer receives and in what order, based on purchase propensity per product, commercial KPIs and history. The goal: sell more and improve other commercial metrics.
Visit time optimization
Led and built
An algorithm that sets how much time to spend with each customer by combining four modules: customer relevance, competitive opportunities, service level and customer independence. It enabled targeted actions with promoters and customers and improved visit times.
Campaign adherence probability
Contributed
A model that estimates how likely each customer is to join a company campaign. It contributed to higher customer participation in campaigns.
Asset allocation by channel
Built
An initiative to measure the benefit of placing commercial assets in different sales channels. That analysis improved how resources are allocated to customers.
Skills
Languages
- Python
- SQL
- PySpark
- TypeScript
- JavaScript
- C# / .NET
- Shell
Data & cloud
- Databricks
- Apache Spark
- Delta Lake
- Apache Airflow
- Azure Data Factory
- Azure Data Lake
- Azure Functions
- Azure Synapse
- Trino / Presto
Machine learning & analytics
- Machine Learning
- Pandas
- NumPy
- Matplotlib
- Seaborn
- Power BI
- Metabase
- Excel
Ways of working
- Scrum
- Kanban
- Azure DevOps
- Git
- Technical leadership
Education
Data Science
Digital House
Feb 2023 – Nov 2023
End-to-end Python projects: exploratory analysis, machine learning models and production deployment.
Data Analytics
Coderhouse
Mar 2022 – May 2022
Relational databases with SQL and Power BI dashboards with DAX, M and Power Query.
Information Systems Engineering
Universidad Tecnológica Nacional
2016 – 2021, through 4th year
Second place at Tecnomate 2016, the faculty’s programming contest. Systems lab intern (2019–2020).
Continuous learning
Platzi
ongoing
Data Science school: descriptive statistics, mathematical functions for AI, Google Cloud Platform, plus English and soft skills.
Contact
Have a hard data or AI problem? Tell me about it.