Vinitra Swamy

Microsoft AI Frameworks · UC Berkeley EECS · Deep Learning Research · vinitra@berkeley.edu

Hello! I am an AI engineer and researcher working on deep learning framework interoperability at Microsoft. I currently work on the Open Neural Network eXchange project (ONNX), an open format to represent deep learning models and translate between AI frameworks.

Before moving to Seattle, I graduated 2 years early from my B.A. and M.S. in Computer Science at UC Berkeley (‘17, ‘18, go bears!) and was a summer lecturer for the Division of Data Sciences.

I love people, data, and working on exciting problems at the intersection of the two:

  • AI for computing education (autograding, performance modeling, scalable infrastructure)
  • Natural Language Processing (textual style transfer, sentiment analysis)
  • Reinforcement Learning reward design / fairness in ML
  • Running a data science course with 1000+ passionate undergrads

Thank you for taking time out of your day to find out what I do with mine!

Research

ONNX: Open Neural Network eXchange

Open Neural Network Exchange (ONNX) is an open ecosystem that empowers AI developers to choose the right tools as their project evolves. ONNX provides an open source format for AI models, both deep learning and traditional ML. It defines an extensible computation graph model, as well as definitions of built-in operators and standard data types. Currently we focus on the capabilities needed for inferencing (scoring).

Microsoft MLADS Conference (Machine Learning, AI, and Data Science)

Gave a talk to data scientists and engineers at MLADS Spring 2019 on model operationalization and acceleration with ONNX alongside Emma Ning, Spandan Tiwari, Nathan Yan, and Lara Haidar-Ahmad.

[Notebooks] [Slides]

University of Washington eScience Institute

Overview of AI model interoperability with ONNX and ONNX Runtime for data scientists and researchers at University of Washington, Seattle.

[Slides]

Conferences attended on behalf of Microsoft AI

Microsoft //Build 2019, KDD 2019, Microsoft Research Faculty Summit 2019, UC Berkeley AI for Social Impact Conference 2018, Women in Cloud Summit 2018, RISECamp 2018

[ONNX + Azure ML Tutorials]

2019

Machine Learning for Humanitarian Data: Tag Prediction using the HXL Standard

Microsoft AI & Research, United Nations OCHA, UC Berkeley

We present a machine learning model to predict tags for datasets from the United Nations Office for the Coordination of Humanitarian Affairs (UN OCHA) with the labels and attributes of the Humanitarian Exchange Language (HXL) Standard for data interoperability. This paper details the methodology used to predict the corresponding tags and attributes for a given dataset with an accuracy of 94% for HXL header tags and an accuracy of 92% for descriptive attributes. Compared to previous work, our workflow provides a 14% accuracy increase and is a novel case study of using ML to enhance humanitarian data.

Collaborated with Elisa Chen, Anish Vankayalapati, Abhay Aggarwal, Chloe Liu (UC Berkeley); Vani Mandava (Microsoft Research); Simon Johnson (UN OCHA). Talk, poster, and short paper published at Social Impact Workshop at KDD 2019 in Anchorage, Alaska.

[Paper] [Talk Abstract] [Slides] [Poster] [Code]

2019

Automating Infrastructure for Data Science Education at Scale

Master's Thesis: Pedagogy, Infrastructure, and Analytics for Data Science Education at Scale

A detailed research report on autograding, analytics, and scaling JupyterHub infrastructure highlighted in use for thousands of students taking Data 8 at UC Berkeley. Presented as a graduate student affiliated with RISELab.

[Thesis]


Deep Knowledge Tracing for Student Code Progression

Knowledge Tracing is a body of learning science literature that seeks to model student knowledge acquisition through their interaction with coursework. This project uses a recurrent neural network (LSTM) to optimize prediction of student performance in large scale computer science classes.

Collaborated with Samuel Lau, Allen Guo, Madeline Wu, Wilton Wu, Professor Zachary Pardos, Professor David Culler on a short paper published at published at Artificial Intelligence in Education / International Festival of Learning 2018 in London, England.

[Paper] [Poster]


Project Jupyter: Scaling JupyterHub architecture for Data 8

Helped develop UC Berkeley data science's software infrastructure stack including JupyterHub, autograding with OkPy, Gradescope, and authentication for 1000s of students.

[Blog] [Code]


Collaborated with Yuvi Panda, Ryan Lovett, Chris Holdgraf, and Gunjan Baid on a talk detailing the infrastructure stack at JupyterCon 2017.

[Slides] [Speaker Profile]

2018

CSi2: Idle Server Identification

IBM Research, T.J Watson Research Center

Recent studies have shown that "zombie" virtual machines in hybrid/private clouds have been wasting millions of dollars worth of resources. The CSi2 algorithm is an ensemble machine learning algorithm to detect inactivity of VMs as well as suggest a course of action like termination or snapshot. It is projected to save IBM Research at least $3.2 million dollars with 95.12% recall and 88% F1 score (>> industry standard) and is being implemented into the Watson Services Platform. 2 patents have been filed and papers are in the process of being finished.

Collaborated with Neeraj Asthana, Sai Zheng, Ivan D'ell Era, Aman Chanana.

2017

Neural Style Transfer for Non-Parallel Text

Natural Language Processing

Expanded on an MIT CSAIL paper by Shen et. al. to improve the accuracy of neural style transfer for unaligned text using author disambiguation algorithms.

Collaborated with Vasilis Oikonomou and Professor David Bamman.

[Code]


2017

Blog Post: Deep Sentiment Analysis

Natural Language Processing

A Distill-style (interactive visualization) introduction and literature review of neural networks for Natural Language Processing, specifically in the field of analyzing sentiment and emotion.

Collaborated with Stefan Palombo and Michael Brenndoerfer.

[Blog]

2017

Deep Causal Reward Design

Fairness in Machine Learning

Exploring reward design for reinforcement learning through the framework of causality and fairness. Class project later expanded into a short paper at CausalML workshop at NeurIPS 2018 by collaborators.

[Paper]

2017

Other Research Projects

Deep DJ: Musical Score Generation for Video

Extracting sentiment from video frames, experimenting with GANs for audio, and ultimately using a neural style transfer for audio technique to generate unique musical tracks for video.

[Blog] [Code]

Goodly Labs: Deciding Force

Advising on data science project to extract key information surrounding police activity from news articles.

[Project Overview] [Website]

BIDS: Ecosystem Mapping Initiative

ETL pipeline and web scraper to determine graph of collaborations between professors and researchers across institutions.

[Code]

2018

Experience

Software Engineer, AI Frameworks

Microsoft

Working on a framework for deep learning / ML framework interoperability (ONNX) alongside an ecosystem of converters, containers, and inference engines.

2018 - Current

Research Assistant

Berkeley Insitute for Data Science (BIDS), RISELab

Worked on projects in AI + Systems with an application area of data science education. Project areas include JupyterHub architecture, custom deployments, OkPy autograding integration, Jupyter noteboook extensions, and D3.js / PlotLy visualizations for data science explorations of funding and enrollment data.

[BIDS] [RISELab]

2015 - 2018

Research Scientist Intern, Machine Learning

IBM Research

Worked on the CSi2 project as a Machine Learning Research Scientist intern on the Hybrid Cloud team. Presented an exit talk and filed 2 patents.

2017

Software Engineering Intern

LinkedIn

Interned at LinkedIn headquarters with the Growth Division's Search Engine Optimization (SEO) Team the summer before entering UC Berkeley. Worked on fullstack testing infrastructure for the public profile pages, as well as a minor Hadoop project; outside of assigned work, helped plan LinkedIn’s DevelopHER Hackathon and worked on several Market Research/User Experience Design initiatives.

2015

CAPE Intern, Made w/ Code Ambassador

Google

Spent a summer learning the fundamentals of Computer Science, algorithms, and computational thinking at Google Headquarters in Mountain View, CA. Chosen as a Google Ambassador for Computer Science following the experience. Worked with Google, Salesforce, and AT&T to introduce coding to over 15,000 girls across California with the Made w/ Code Initiative.

2011

Education

University of California, Berkeley

Master's in Electrical Engineering and Computer Science
  • President of Computer Science Honor Society (UPE)
  • Head Graduate Student Instructor of Data 8 (Foundations of Data Science)
  • Graduated at 20 as youngest graduate of the M.S. in EECS program in recent UC Berkeley history
  • Advisor: Dean of Data Sciences, Freisen Professor in EECS, David Culler
2017 - 2018

Teaching: Foundations of Data Science

UC Berkeley Data 8
  • Summer 2018 Lecturer and Head GSI of Data 8 for 3 semesters, responsible for management of 1000+ undergraduates, 40 GSIs, 30 tutors, and 100+ lab assistants each semester.
  • Helped create data science curriculum material for lecture and domain-specific seminar courses.
  • In charge of maintaining JupyterHub infrastructure for 1500+ active users (with Docker/Kubernetes backend on top of Google Cloud, Azure, and AWS).

[Data 8 Website] [Course Offering] [Code]

2016 - 2018

University of California, Berkeley

Bachelor's in Computer Science
  • EECS Award of Excellence in Undergraduate Teaching and Leadership
2015 - 2017

Speaking Engagements and Awards

Speaking Engagements

  • Summer 2019: Guest on the Microsoft AI Show
  • Spring 2019: Speaker at Microsoft Machine Learning and Data Science Conference (MLADS)
  • Summer 2018: Lecturer for UC Berkeley's Data 8, alongside Fahad Kamran (250+ students)
  • Summer 2018: Presenter at Artificial Intelligence in Education 2018 in London, England
  • Summer 2018: Speaker at UC Berkeley's Data Science Undergraduate Pedagogy and Practice Workshop
  • Fall 2017: Opening Panelist at SalesForce Dreamforce
  • Summer 2017: Speaker at JupyterCon NYC
  • Spring 2017: Presenter at Berkeley Institute for Data Science Research Showcase
  • Fall 2016: Panelist at SF BusinessWeek Conference
  • Summer 2016: Conference organizing team at Algorithms for Modern Massive Data Sets (MMDS)

Awards

  • UC Berkeley EECS Award of Excellence for Teaching and Leadership
  • Google International Trailblazer in Computer Science
  • UC Berkeley Graduate Opportunity Fellowship Recipient
  • Kairos Fellow, UC Berkeley
  • Inducted into UPE, UC Berkeley Computer Science Honor Society
  • UC Berkeley Alumni Leadership Scholar
  • NASA-Conrad Foundation Spirit of Innovation Cybertechnology Finalist
  • Girl Scout Gold Award: Bridging the Digital Divide