Aram Ebtekar, Developer in Vancouver, BC, Canada
Aram is available for hire
Hire Aram

Aram Ebtekar

Verified Expert  in Engineering

Artificial Intelligence Developer

Location
Vancouver, BC, Canada
Toptal Member Since
February 11, 2022

Aram is an open-source Rust developer, multi-disciplinary researcher, and applied mathematician with graduate-level training in computer science, mathematics, physics, and economics. He has a competitive programmer, coach, and judge background, placing himself on the world's top 70 in the Google Code Jam and Topcoder Open. Presently, Aram is the fourth member of Waymo's behavior prediction team.

Portfolio

Independent Work
Rust, Mathematics, Physics, Information Theory, Algorithms, Game Theory
Mythic
Python 3, Deep Learning, Computer Vision
Waymo LLC
Applied Mathematics, Algorithms, Data Structures, C++17

Experience

Availability

Part-time

Preferred Environment

Rust, Self-driving Cars, Information Theory, Algorithms, Applied Mathematics, Data Structures, Reinforcement Learning, Linux

The most amazing...

...open-source I've delivered is the Rust algorithms repository, launched at #1 daily trending on GitHub and #2 on Hacker News.

Work Experience

Self-directed Advanced Studies and Research

2019 - PRESENT
Independent Work
  • Developed the Elo-MMR skill estimation algorithm, which is receiving considerable interest from industry, hobbyists, and academia. While classic Elo is suited for two-player games, Elo-MMR is the first principled system for large ranked competitions.
  • Built a mathematical model for emergent causality, suggesting a possible solution to the problem of the perceptual arrow of time in physics.
  • Acquired graduate-level expertise and problem-solving intuition in a variety of additional computer science topics, as well as in parts of mathematics, physics, and economics.
Technologies: Rust, Mathematics, Physics, Information Theory, Algorithms, Game Theory

Senior AI Research Scientist

2018 - 2019
Mythic
  • Led AI co-design research efforts to understand how efficiently different types of convolutional layers map onto hardware and retain the model's accuracy over time.
  • Led the development of a video quality enhancer by modifying academic deep learning models to suit practical datasets and constraints.
  • Contributed to technical writing and explanatory documentation on various topics, thus enabling interdisciplinary communication and collaboration.
  • Owned the company's algorithmic whiteboard interviews for job candidates, leveraging my experience as a competitive programming contestant and problem setter.
Technologies: Python 3, Deep Learning, Computer Vision

Behavior Prediction Researcher and Engineer

2016 - 2018
Waymo LLC
  • Invented and deployed behavior prediction algorithms on self-driving cars, achieving state-of-the-art speed and robustness in predicting other drivers' movements.
  • Coordinated and taught two machine learning courses to engineers in weekly sessions with assigned readings and university course materials.
  • Advocated first-principles and data-driven approaches through teaching and documenting how my work resulted in simpler algorithms and software architecture.
Technologies: Applied Mathematics, Algorithms, Data Structures, C++17

Elo-MMR Rating System

https://github.com/EbTech/Elo-MMR
A rating system for tracking and ranking players' skills in competitions.

I was the algorithm's inventor and the principal researcher and engineer of the project. We released open-source software to rank custom datasets and benchmark different rating systems. Now we're in the process of deploying a live presentation of ratings for multiple sports at worldrank.org

Contest Algorithms in Rust

https://github.com/EbTech/rust-algorithms
A collection of idiomatic algorithm and data structure implementations in Rust, suitable for educational and competitive programming use. Perhaps for the first time, it proved that the Rust language is also suitable for writing short and simple programs. It has about 3,000 GitHub stars.

Information Dynamics and the Arrow of Time

https://arxiv.org/abs/2109.09709
A solution to the perceptual arrow of time problem. We proposed a stochastic partitioned cellular automaton (SPCA) model to study the mechanisms and consequences of emergent irreversibility, showing that SPCA dynamics can be deterministic and reversible by attaching randomly initialized degrees of freedom.

Languages

Rust, C++17, Python 3, Java, Haskell, C#, C, Prolog

Other

Information Theory, Algorithms, Applied Mathematics, Data Structures, Artificial Intelligence (AI), Causal Inference, Technical Writing, Research, Self-driving Cars, Reinforcement Learning, Robotics, Formal Methods, Physics, Game Theory, Machine Learning, Concurrency, Economics, Mathematics, Deep Learning, Computer Vision, Distributed Systems

Libraries/APIs

Keras, PyTorch

Platforms

Linux

Storage

Databases

2012 - 2015

Master's Degree (Within a PhD Program) in Computer Science

Carnegie Mellon University - Pittsburgh, PA, USA

2008 - 2012

Bachelor's Degree in Mathematics and Computer Science

University of British Columia - Vancouver, BC, Canada

Collaboration That Works

How to Work with Toptal

Toptal matches you directly with global industry experts from our network in hours—not weeks or months.

1

Share your needs

Discuss your requirements and refine your scope in a call with a Toptal domain expert.
2

Choose your talent

Get a short list of expertly matched talent within 24 hours to review, interview, and choose from.
3

Start your risk-free talent trial

Work with your chosen talent on a trial basis for up to two weeks. Pay only if you decide to hire them.

Top talent is in high demand.

Start hiring