I’m a software engineer specializing in artificial intelligence and machine learning, with a background in computational mechanics and spacecraft engineering. I'm data driven, passionate about learning, and strive to build a better world with technology.

 

Work Experience

Senior Research Engineer
Vicarious AI
October 2016 +
vicarious.com

We're building general artificial intelligence. My focuses:

  • Project lead for developing robotic motion planning and trajectory optimzation algorithms based on Gaussian processes
  • Building probabilistic graphical models (PGM) for robotic vision, with inspiration from primate visual cortex
  • Lead software architect for mono-repo of Python, C++, and ROS code
Senior Software & Research Engineer
Numenta
November 2014 - October 2016
numenta.com/learnnumenta.org

Numenta aims to reverse-engineer the neocortex for machine intelligence. My role was mainly to prototype algorithms from raw theory, test and validate them through experiments, and develop them into production quality code.

  • Experimented with algorithms for natural language processing (NLP), and managed the development of a document classification application
  • Investigated and implemented algorithms to compare with HTM -- e.g., deep learning (LSTM), traditional time-series and Bayesian methods, word embeddings (word2vec) -- for sequence prediction, classification, and anomaly detection
  • Built and managed the open-source Numenta Anomaly Benchmark
  • Helped architect CI pipelines for using Docker
  • Community leader, including numerous public speaking engagements at conferences, meetups, and workshops
  • Key role in all publishing, including writing chapters for our textbook "BaMI"
  • https://discourse.numenta.org/u/alavin/summary
Research Associate
NASA Ames Research Center
Summer 2013

Applied new approach to designing a re-entry system for on-demand return of scientific payloads from the International Space Station.

Data Analytics Specialist
Northwestern University and Agentis Energy
June - December 2012

Provided energy company with a new approach (k-NN clustering) to better classify customers based on true energy usage patterns and investigate energy efficiency traits (see Publications below).

Rocket Propulsion Intern
Technion Institute of Technology
Summer 2013

Designed experiments and testbed to investigate the use of nitrous oxide in hybrid propulsion engines. Ran rocket engine firing tests and analyzed results across a variety of performance metrics.

 

Education

M.S.
Mechanical Engineering
Carnegie Mellon University
Class of 2014
  • computational mechanics
  • mobile robotics
  • artificial intelligence & machine learning
  • engineering computation
GPA: 4.0
Master of
Engineering Management
Duke University
2013 - never
  • software development management
  • commercializing technology innovation

I dropped out. Why? I found most of the curriculum unfulfilling, and dedicating myself to learning engineering management skills through experience to be far more valuable.

GPA: 3.9
  • Concentration in Technology Innovation
B.S.
Mechanical Engineering
Cornell University
Class of 2012
  • spacecraft engineering
  • mechatronics
GPA: 3.1 (3.6 in '11-'12)
  • Satellite dynamics research in Space Systems Design Studio
  • College of Engineering Dean's List, 2011-2012
  • Two-time Global Fellow for engineering work in France and Israel
  • Men's varsity golf team, 4-year starter and team captain

 

Technical Skills

Main languages: Python and C++. Main tools: NumPy, Eigen, gtsam, git, ROS. Ping me for more granularity.

Programming

Expert
  • Python
  • C++
  • Julia
  • Java
  • MATLAB
  • Go
  • Android
  • R

Data Sci and ML Tools

Advanced
  • keras
  • scikit-learn
  • tensorflow
  • pandas
  • numpy
  • plotly
  • nltk
  • opencv
  • edward
  • pyro
  • eigen
  • ceras
  • gtsam
  • openai gym

SW, etc.

  • architecture
  • Docker
  • GitHub API
  • AWS
  • continuous integration
  • git
  • web apps
  • html
  • raspberry Pi
  • ros
  • robotics simulators/engines

 

Awards

"Lavin is something of a polymath, having been a literal rocket scientist at NASA and Blue Origin before turning his hand to AI development."

2016
Cornell U., College of Engineering Dean's List
2011 & 2012
Cornell U., Engineering Global Fellow
2010 & 2011

 

Additional Projects

2016
Deep Learning for Self-Driving Vehicles
GitHub repo

Implemented varieties of CNN + LSTM architectures to investigate temporal dependencies in video data, specifically for predicting speed and steering for a self-driving car.

  • Implemented a scaled down version of NVIDIA's "End to End Learning for Self-Driving Cars" model.
  • Experimented with different feature pooling architectures as means of incorporating temporal info into CNNs.
  • Experimented using optical flow tracking to improve model accuracy.
2014
Mobile Robotics Path Planning Algorithms

Integrated Pareto-optimality into traditional A* and D* search algorithms to provide improved methods for multi-objective path planning. I showed strong improvements over traditional methods across all optimization metrics in simulated environments, including in lunar and Mars terrain case studies.

2013-2014
Google Lunar XPrize
GLXP team Astrobotic

Lunar rover project manager and lead of systems engineering.

2014
Kaggle Higgs Boson Machine Learning Challenge
Higgs Challenge

For the task of classifying particle collision events as either signal (Higgs boson decay) or background noise, I implemented a gradient boosted classifier that performed only 4.63% behind the winning score, and almost 2x the score of the baseline naive Bayes model.

2013-2014
Carnegie Mellon Master's Thesis

"Finite Element-based Structural Optimization of Large System Models Under Buckling Constraints"

  • Identified the core issues with applying standard optimization methods to the problem, and wrote a new algorithm to address convergence issues and computational efficiency.
  • arXiv paper and thesis presentation
2011-2012
Spacecraft Research at Cornell
  • Designed and built attitude control system for a multiple CubeSat configuration; utilized DC motor inertia in place of reaction wheels, saving on mass, power, and spending budgets.
  • Designed the allocation and layouts of CubeSat payloads for efficient dynamics.

 

Publications

Note there are additional papers linked in the above projects section. Also see my Google Scholar and arXiv profiles.

Science, 2017

Drawing inspiration from systems neuroscience, we introduce a probabilistic generative model for vision in which message-passing based inference handles recognition, segmentation, and reasoning in a unified way. The Recursive Cortical Network (RCN) outperforms state-of-the-art deep learning models on challenging text recognition benchmarks, while being 300x more data efficient. See our "Common sense, cortex, and CAPTCHA" blog post for more. Also, I was lead dev for the reference implementation.

ProbProg 2018

Abstract: Many probabilistic programming languages decouple modeling and inference: the user specifies a model, and inference is delegated to an inference engine based on black-box MCMC or variational techniques. This approach has two limitations that inhibit its use in robotics. First, many robotics applications require real-time performance. Inference algorithms thus need to be optimized for each application. Second, many applications of robotics learning require combining probabilistic inference with optimization, in settings where evaluating the objective function requires costly real-world interactions. Modeling and inference must thus be integrated with Bayesian optimization techniques. We propose a design for a domain-specific probabilistic programming language for data-efficient robotics. The design prominently features (i) programmable inference, (ii) modular and structured optimization, and (iii) explicit uncertainty quantification. We argue that such a DSL could be useful for tasks such as Bayesian optimization for grasping, and learning policies for control.

CCN 2018

Abstract: Understanding the information processing roles of cortical circuits is an outstanding problem in neuroscience and artificial intelligence. The theoretical setting of Bayesian inference has been suggested as a framework for understanding cortical computation. Based on a recently published generative model for visual inference (George et al., 2017), we derive a family of anatomically instantiated and functional cortical circuit models. In contrast to simplistic models of Bayesian inference, the underlying generative model's representational choices are validated with real-world tasks that required efficient inference and strong generalization. The cortical circuit model is derived by systematically comparing the computational requirements of this model with known anatomical constraints. The derived model suggests precise functional roles for the feedforward, feedback and lateral connections observed in different laminae and columns, and assigns a computational role for the path through the thalamus.

CCN 2018

Abstract: The connectivity and information pathways of visual cortex are well studied, as are observed physiological phenomena, yet a cohesive model for explaining visual cortex processes remains an open problem. For a comprehensive understanding, we need to build models of the visual cortex that are capable of robust real-world performance, while also being able to explain psychophysical and physiological observations. To this end, we demonstrate how the Recursive Cortical Network (George et al., 2017) can be used as a computational model to reproduce and explain subjective contours, neon color spreading, occlusion vs. deletion, and the border-ownership competition phenomena observed in the visual cortex.

"A Computational Model of Ventral & Dorsal Processing"
(in-progress)

Models of neocortex largely consider the ventral pathway of the vision system, yet integration with the parallel dorsal pathway is necessary for sensorimotor processing. I present a ventral + dorsal computational model with an updated canonical microcircuit.

Neurocomputing, 2017

Propose a novel anomaly detection algorithm based on Hierarchical Temporal Memory (HTM), an online sequence memory algorithm derive from the neocortex.

ebook, 2016

An online textbook on Hierarchical Temporal Memory with Numenta.

ICMLA, 2015

Paper and oral presentation at 2015 IEEE International Conference on Machine Learning Applications

Energy Efficiency, 2014

Journal paper in Energy Efficiency

 

Interests

Reading

  • Heinlein
  • Game of Thrones
  • spacetime and string theory
  • philosophy of physics
  • Dawkins

Health & fitness

  • half marathons
  • yoga
  • keto/paleo
  • meditation