Jacob Danovitch

Summary

  • MSc student at MILA & McGill University, supervised by Reihaneh Rabbany.
  • Data science intern at Microsoft Search, Assistance, and Intelligence.
  • Developing methods for search and classification on heterogeneous, structured data.
  • Particularly interested in social good applications.

Education

Montreal Institute for Learning Algorithms

Montreal, Canada

MSc Machine Learning | Cum. GPA: 4
  • Supervised by Professor Reihaneh Rabbany at McGill University.
  • Working on fact verification, human trafficking detection, citation recommendation.
  • Assisting with web scraping and infrastructure management.

Carleton University

Ottawa, Canada

BCS Computer Science, Honors | Minor in Psychology | Cum. GPA: 4
  • Graduated August 2020 with high distinction.
  • Deans' Honour List from 2018-2020.
  • Completed undergraduate thesis supervised by Professor Majid Komeili.

Experience

Microsoft Search, Assistance, and Intelligence

Seattle, USA

Data Science Intern
  • Training personalized model for email search with Microsoft Outlook Search Relevance team.
  • Implemented distributed training pipeline in eyes-off environment using PySpark, PyTorch, and Azure ML.
  • Received return offer for data scientist internship beginning May 2021.

Carleton University

Ottawa, Canada

Teaching Assistant
  • Fall 2019: Assisted Professor David McKie with JOUR4401: Storytelling with Data.
  • Gave several lectures on tools for web scraping and data science, assisted journalism students in-class.
  • Previously: Web Development, Discrete Structures I, Introduction to Computer Science I.

Microsoft Search, Assistance, and Intelligence

Seattle, USA

Software Engineering Intern
  • Developed character-RNNs for classification and NER to extract meeting details from Microsoft Outlook emails.
  • Deployed models for inference and integrated into Outlook's React frontend.
  • Patent pending for contributions to Contextual Suggestions project.

Microsoft Cortana

Seattle, USA

Explore Intern
  • Gained engineering and project management experience as Explore intern.
  • Contributed to Cortana-enabled data visualization and feedback collection web application.
  • Implemented efficient backend with Azure-SQL and user-facing dashboard using C# and .NET.

Medimo Labs

Toronto, Canada

Software Development Intern
  • Contributed to medical data visualization application.
  • Implemented client-facing dashboard and search functionality with React.js.
January 30, 2021 Jacob Danovitch · Résumé 1

Short-Term

Allen Instititute for AI

Virtual

VIsitor
  • Working with the internal AllenNLP team to improve distributed training.
  • Implementing ZeRO optimizer using Microsoft Deepspeed to scale language model training.
  • Early benchmarks show >50% reduction in GPU memory usage.

Canadian Broadcasting Corporation (CBC)

Ottawa, Canada

Data Science Consultant
  • Scraping, cleaning, and aggregating data on health violations in long-term care homes.
  • Implemented high-performance scraping pipeline to parse semi-structured PDFs from government reports.

Research

Publications

Pelrine, K* and Danovitch, J* and Rabbany, R. (2021). The Surprising Performance of Simple Baselines for Misinformation Detection. Proceedings of The Web Conference 2021.

Pelrine, K* and Danovitch, J* and Orozco Camacho, A and Rabbany, R. (2020). Detecting Informative COVID-19 Tweets by Attending over Linked Documents. Proceedings of the 6th Workshop on Noisy User-generated Text.

Danovitch, J. (2020). Do Human-Written Explanations Make Neural Attention More Interpretable?. Undergraduate thesis.

Danovitch, J. (2020). Linking Social Media Posts to News with Siamese Transformers. 9th International Conference on Information Technology Convergence and Services (ITCSE 2020), Vancouver, CA.

Danovitch, J. (2019). Trouble with the Curve: Predicting Future MLB Players Using Scouting Reports. 2019 Carnegie Mellon Sports Analytics Conference, Pittsburgh, USA.

Lynch, B., Danovitch, J., & Davies, J. (2018). Towards a Computational Approach to Conceptual Metaphor. Poster session at CogSci 2019, Montreal, CA.

Patents

Fenton et al. (2019). Composing Rich Content Messages. (MS# 406941-US-NP). U.S. patent application.

Grants & Awards

NSERC CGS-M Research Award

$17500 (declined)

Carleton University

Clarence C. Gibson Scholarship

$3000

Carleton University

Reproducible Research Competition Finalist

$1000

Carnegie Mellon Sports Analytics Conference

I-CUREUS Grant

$2250

Carleton University

E.W.R. Steacie Scholarship

$3000

Carleton University

Lester B. Pearson Scholarship

$3000

Carleton University

President's Scholarship

$3000

Carleton University