Pedro T.

Staff Machine Learning Scientist at Intercom
  • Claim this Profile
Contact Information
us****@****om
(386) 825-5501
Location
Ireland, IE
Languages
  • Portuguese Native or bilingual proficiency
  • English Full professional proficiency
  • Spanish Elementary proficiency

Topline Score

Topline score feature will be out soon.

Bio

Generated by
Topline AI

5.0

/5.0
/ Based on 1 ratings
  • (1)
  • (0)
  • (0)
  • (0)
  • (0)

Filter reviews by:

LinkedIn User

Trabalhei com o Pedro Tabacof no time de Data & Analytics do iFood. Tive o prazer de contribuir com parte da disponibilidade dos dados usados nos projetos de machine learning que ele conduziu. Pedro Tabacof é muito inteligente e sabe ser crítico de uma forma bastante construtiva. Sou grato por tudo que aprendi trabalhando com ele.

You need to have a working account to view this content.
Join now
You need to have a working account to view this content.
Join now

Credentials

  • Machine Learning Data Lifecycle in Production
    Coursera
    Dec, 2021
    - Nov, 2024
  • Introduction to Machine Learning in Production
    Coursera
    Nov, 2021
    - Nov, 2024
  • Inferential Statistics
    Coursera
    Jun, 2017
    - Nov, 2024
  • Introduction to Probability and Data with R
    Coursera
    Jan, 2017
    - Nov, 2024
  • Functional Program Design in Scala
    Coursera
    Sep, 2016
    - Nov, 2024
  • Deep Learning Summer School
    Université de Montréal
    Aug, 2016
    - Nov, 2024
  • Functional Programming Principles in Scala
    Coursera
    Aug, 2016
    - Nov, 2024
  • Machine Learning Summer School
    The University of Texas at Austin
    Jan, 2015
    - Nov, 2024
  • Six Sigma - Green Belt
    EDTI - Analise de Dados e Melhoria Organizacional
    Nov, 2014
    - Nov, 2024

Experience

  • Intercom
    • United States
    • Software Development
    • 700 & Above Employee
    • Staff Machine Learning Scientist
      • Apr 2023 - Present

      NLP. Chatbots. Real-time models. Python. AWS. Snowflake. DBT. NLP. Chatbots. Real-time models. Python. AWS. Snowflake. DBT.

  • Wildlife Studios
    • Brazil
    • Mobile Gaming Apps
    • 700 & Above Employee
    • Data Science Manager
      • Mar 2022 - Apr 2023

      Manager of 3 senior data scientists working with marketing attribution and lifetime value. Promoted a senior DS to staff.Mentored more than 5 newhires during their month long onboarding. Created and taught learning tracks on PySpark, Databricks and Airflow to multiple data scientists.

    • Staff Data Scientist
      • Feb 2021 - Apr 2023

      Marketing attribution in a privacy-aware world. Led the design and implementation of a solution to probabilistically attribute users to campaigns without personal identifiers, only using Apple's SKAN. Implemented Bayesian inference using PySpark to handle the scale. All iOS performance marketing decisions were based on the outputs of this system after Apple's tracking transparency change in 2021 (performance marketing budget of ~10M USD / year). Led the data science aspects of a taskforce that included the head of UA and the CTO. Show less

    • Senior Data Scientist
      • Dec 2019 - Feb 2021

      Lifetime value modeling for performance marketing. Designed LTV ML models for two new games that were huge hits (Zooba and Tennis Clash, 100M+ downloads). Improved LTV forecasting using Kalman Filters and ML models for those games and more. Significant impact to the bottomline (personal ROI of at least 10x). Used PySpark and LightGBM.

  • Accenture
    • Ireland
    • Business Consulting and Services
    • 700 & Above Employee
    • Research Engineer
      • Jun 2019 - Dec 2019

      **ICLR 2020 publication: Probability Calibration for Knowledge Graph Embedding Models.** Software engineering for AI research. Knowledge graph embeddings, counterfactual explanations, and Bayesian deep learning. Contributor to the TensorFlow-based AmpliGraph library. Part of the winning team of the DatSci 2019 awards (best technical advance in the field of data science / AI from a research organization). **ICLR 2020 publication: Probability Calibration for Knowledge Graph Embedding Models.** Software engineering for AI research. Knowledge graph embeddings, counterfactual explanations, and Bayesian deep learning. Contributor to the TensorFlow-based AmpliGraph library. Part of the winning team of the DatSci 2019 awards (best technical advance in the field of data science / AI from a research organization).

  • Nubank
    • Brazil
    • Financial Services
    • 700 & Above Employee
    • Senior Data Scientist
      • May 2018 - Apr 2019

      Risk and severity modeling using the latest machine learning Python tech stack (XGBoost, LightGBM, scikit-learn). Deployed models that had a significant impact on the business (credit lines). Policy and Net Present Value (NPV) framework modeling, bridging the gap from models to decisions. ETL and data engineering in Spark / Scala. Core contributor to the now open source fklearn library (functional machine learning in Python). Risk and severity modeling using the latest machine learning Python tech stack (XGBoost, LightGBM, scikit-learn). Deployed models that had a significant impact on the business (credit lines). Policy and Net Present Value (NPV) framework modeling, bridging the gap from models to decisions. ETL and data engineering in Spark / Scala. Core contributor to the now open source fklearn library (functional machine learning in Python).

  • iFood
    • Brazil
    • Software Development
    • 700 & Above Employee
    • Senior Data Scientist
      • Oct 2017 - May 2018

      Predictive modeling for antifraud system. Millions of rows, hundreds of features, a lot of money saved. Spark and XGBoost for ETL and modeling. Deployment using Scala. Predictive modeling for antifraud system. Millions of rows, hundreds of features, a lot of money saved. Spark and XGBoost for ETL and modeling. Deployment using Scala.

  • Motorola Mobility (a Lenovo Company)
    • United States
    • Computers and Electronics Manufacturing
    • 700 & Above Employee
    • Researcher (part-time)
      • Feb 2017 - Oct 2017

      Applied predictive modeling for smartphone bug detection and classification. Information retrieval. Natural language processing for feature extraction. Development of web service to serve predictions and provide interpretability. Applied predictive modeling for smartphone bug detection and classification. Information retrieval. Natural language processing for feature extraction. Development of web service to serve predictions and provide interpretability.

  • Datart
    • Brazil
    • IT Services and IT Consulting
    • 1 - 100 Employee
    • Director of Research
      • Sep 2016 - Oct 2017

      Development of Call.AI, web application for speech analytics and audio retrieval. Data science consulting work in predictive modeling for e-commerces, including prediction of conversion rates, churn, and customer lifetime value. Development of Call.AI, web application for speech analytics and audio retrieval. Data science consulting work in predictive modeling for e-commerces, including prediction of conversion rates, churn, and customer lifetime value.

  • iSystems
    • Brazil
    • Software Development
    • 1 - 100 Employee
    • Research and Development Engineer
      • Aug 2013 - Sep 2016

      * R&D engineering for advanced process control software. * C++ development of industrial desktop applications.* Creation of stochastic optimization and sensor fusion modules.* Improvements on Fuzzy and classical control modules.* Full-stack development: Core library (STL based), application and GUI (Qt).* Python (IPython + Spyder) and R (RStudio) for prototyping.* Management of research internships.

    • Intern
      • Aug 2012 - Jul 2013

      Worked on a research project on nonlinear modelling of dynamical systems using a grey-box approach. Implemented parameter estimation algorithms using nonlinear control theory (single-shooting, direct collocation), meta-heuristics (evolutionary strategies) and neural networks (echo state networks, extreme learning machines).Development done in C/C++ and Python using Eigen, MPFIT, CasADi and Ipopt libraries.

  • Czech Technical University in Prague
    • Czechia
    • Higher Education
    • 700 & Above Employee
    • Lab intern
      • Dec 2011 - Feb 2012

      Photovoltaic (PV) simulation through numerical computing and SPICE; GUI development using Tkinter and Python. Photovoltaic (PV) simulation through numerical computing and SPICE; GUI development using Tkinter and Python.

  • Universidade Estadual de Campinas
    • Brazil
    • Research Services
    • 700 & Above Employee
    • Teaching assistant
      • Jul 2009 - Jun 2010

      TA for two introductory Computer Science classes, under supervision of Prof. Rogério Drummond (2009) and Prof. Hans Liesenberg (2010), at the Institute of Computing (IC). Assignments and projects were done using the C language. TA for two introductory Computer Science classes, under supervision of Prof. Rogério Drummond (2009) and Prof. Hans Liesenberg (2010), at the Institute of Computing (IC). Assignments and projects were done using the C language.

Education

  • Universidade Estadual de Campinas
    Master of Science (M.Sc.), Electrical Engineering
    2015 - 2017
  • Universidade Estadual de Campinas
    Computer Engineering, 8th out 110 students
    2008 - 2013
  • Bishop's University
    Bachelor's degree, Computer Science
    2010 - 2010

Community

You need to have a working account to view this content. Click here to join now