Skills and Interests
Interested in: Natural Language Processing, Reinforcement Learning, Machine Learning, Artificial Intelligence, Computer Vision, AI Ethics, Reproducibility in Research, and so much more…
Coursework taken (as a student): Reinforcement Learning, Machine Learning, Human Neuroimaging Methods, and more
Languages: English (Native), Russian (Rusty Native), Spanish (Intermediate)
Tech I use often or know well: Pytorch, Tensorflow, Keras, Git, MongoDB, node.js, all the AWS things (DynamoDB, EC2, S3), and more.
Certifications I’m current in: First Aid/CPR, EMT-B (NREMT, MA)
Reviewer for: IROS 2017, IROS 2018, JMLR, AWS ML Research Awards Proposals, Amazon Machine Learning Conference (AMLC) 2018, RML@ICML 2018 Workshop, CORL 2018, ICLR 2019, AAAI 2019 Workshop on Reproducible AI
Organizer of: ICML 2018 Workshop on Reproducibility in Machine Learning
Program Committee Member of: AAAI 2019 Workshop on Reproducible AI
Teacher for: Math 363: Discrete Mathematics (TA, McGill University), Comp 202: Intro to Computing (TA, McGill University), Comp 303: Software Design (TA, McGill University), EMED 111A: Emergency Medical Technician Training (TA, Stanford University)
- Ph.D. Computer Science
Advisor: Dan Jurafsky
- M.Sc. Computer Science
McGill University / Montréal Institute for Learning Algorithms
Thesis: Reproducibility and Reusability in Deep Reinforcement Learning
Advisors: David Meger and Joelle Pineau
- B.Eng. Software Engineering [Minor in Biomedical Engineering]
Thesis: Autonomous Swarm Behaviour in Mesh Networked Agents
Advisor: Mark Coates
- Oct 2017 - July 2018
@ Amazon (Alexa Brain)
Python, PyTorch, Keras, Theano, Tensorflow
- Working on the Alexa Brain Conversational AI team. Making Alexa have better conversations using deep reinforcement learning. (PI: Alborz Geramifard)
- Implemented various RL algorithms to train policies to ask clarifying questions under SLU noise.
- Worked on integration of policies into the MovieBot Alexa Skill
- Sep 2016 - Oct 2017
@ McGill University / Montreal Institute of Learning Algorithms
@ Mobile Robotics Lab / Reasoning and Learning Lab
Python, TensorFlow, MuJoCo, PyTorch
- Conducted research in various aspects of machine learning and reinforcement learning
- computer vision in for robotic systems
- learning hierarchical reward functions and temporal abstraction in IRL
- reproducibility in RL
- ethics in dialogue systems
- and much more.
- Jun 2015 - Sep 2016
Software Development Engineer
@ Amazon (AWS)
DynamoDB, EC2, SimpleDB, S3, EBS, Java, Python, Bash
- Worked on AWS New Initiatives including Storage Gateway and File Gateway.
- Developed internal tool for automatically analyzing high severity events for quicker response times and recommended actions to mitigate the problem.
- Developed tooling for migrating database systems safely and with minimal downtime, used the tools in a live production environment.
- Developed and architected the core code of File Gateway along with a small agile team.
- Supported customers and participated in maintaining the live production Storage Gateway system.
- May 2014 - Aug 2014
Software Engineering Intern
@ A Thinking Ape
- Worked on Realtime Bidding System for buying mobile ad space, initially with Mopub and OpenX
- Used Django/Python, MySQL, Apache, HAProxy, Monit and other server technologies to achieve a round trip time of < 100 ms for tens of thousands of requests per second
- Was the lead developer for a logging, analysis, and machine learning tool for analyzing and learning bidding models using Python, Spark, and Hadoop among other tools. Also created a web interface for ease-of-use and handled initial server configurations, setup, and monitoring.
- Developed additional internal metrics tools for analyzing acquired user value and retention statistics within the company’s mobile games
- Saw Realtime Bidding system to a soft launch for internal advertising
- Sep 2013 - Nov 2013
Research Assistant and Software Developer
@ McGill Centre For BioInformatics
- Worked briefly on Phylo, an online game to crowd source the matching of phylogenetic data
- Jan 2013 - Aug 2013
Research and Software Development Intern
- Worked on Web Communication Gateway, a Java backend application for exposing Real Time Communication services (RCS) to clients over a REST interface that can handle millions of users per rack and hundreds of thousands of users per instance.
- Initially developed and designed proprietary TTCN-3 Framework for SIP/HTTP/MSRP signaling and set up automated testing system through Jenkins
- Quickly became part of the core Research & Development Team developing the commercial product in Java, cross validating and implementing functionality/tests based on RCS/Telecommunications specifications including SIP/MSRP/HTTP signaling and multithreaded systems.
- Saw product from pre-commercial redesign stage to commercial launch
- Sep 2012-Dec 2012
Research Assistant and Software Developer
@ McGill DDMAL Music Technology Laboratory
Python, Django, C++
- Worked on web application pipeline for image processing and Optical Music Recognition in Python called Rodan
Kind of Irrelevant Experience
- Jul 2010 - Aug 2012
- Produce Manager
@ Fancy's Marketplace (Osterville, MA)
- Work seasonally as stock staff. Promoted to Produce Manager Summer, 2012.
- Jun 2009 - Sep 2009
@ Wimpy's Restaurant & Tavern (Osterville, MA)
- Jul 2008 - Sep 2008
- Prep Cook (ServSafe Certified)
@ Dockside Restaurant & Bar (Hyannis, MA)
- Jun 2007 - Sep 2007
- Prep Cook and Cashier
@ Cape Cod Bagel Co. (Falmouth, MA)