Aran Khanna

MLConf Atlanta Keynote: High Performance Deep Learning on Edge Devices

A talk I was invited to give the keynote at MLConf in Atlanta on how using AWS services and MXNet's specialized features for deploying deep learning on IoT, mobile and embedded devices can provide new flexibility as well as cost savings for computer vision algorithms by moving them out of the cloud and onto edge devices

9/15/2017

AWS re:Invent 2017: Deep Learning for Industrial IoT

A talk I was co-presented with NVIDIA on applying deep learning models in the setting of industrial IoT devices, as well as how to utilize AWS services to build production systems around these models.

11/30/2017

ICML 2017: Distributed Deep Learning with MxNet Gluon

A workshop co-presented with professor Alex Smola at ICML 2017 in Sydney demonstrating the Gluon machine learning framework created by our research team at AWS deep learning.

8/6/2017

Software Engineering Daily Podcast: Edge Deep Learning

An invited interview I gave to the Software Engineering Daily podcast diving into the technicals of deep learning and how developers can think about effectively introducing this technology into IoT and Edge systems.

1/26/2018

High-Performance Deep Learning On Embedded Devices

A technical workshop I was invited to co-present with a collaborator from NVIDIA's at their GTC conference, covering how the Apache MXNet framework can be deployed across a variety of IoT devices, such as the NVIDIA Jetson TX2 platform and how AWS services can be used to manage, coordinate and monitor deep learning models running on these devices.

5/8/2017

AWS re:Invent 2017: AWS DeepLens Workshop Building Computer Vision Applications

A technical workshop I co-presented at AWS re:invent 2017 on how to use the features of the newly released DeepLens developer device to build computer vision models in the AWS cloud and deploy to them to run locally, in real-time on the camera.

11/29/2017

Amazon AI Conclave: Running Deep Learning Models on Client Devices

A talk at the Amazon AI Conclave in Bangalore on how to leverage AWS services to build deep learning based computer vision systems that don't rely on constant connection the the cloud, and the benefit of these types of systems in the Indian market.

12/15/2017

AWS re:Invent 2017: Accelerating Apache MXNet Models on Apple Platforms Using CoreML

A technical workshop I co-presented at AWS re:invent 2017 on creating deep learning models in the AWS cloud using Apache MXNet then building a system to optimize these models via Apple's CoreML framework for prediction on iOS and OSX devices.

11/29/2017

AWS Summit Berlin: How to Build a Deep Learning Model That Works on Edge Devices

A technical demo developed and co-presented with a collaborator from NVIDIA at the AWS Berlin Summit showing a deep network based computer vision models running in real time on a Rasberry Pi 3 and an NVIDIA Jetson TX2 coordinated with AWS's IoT service

5/18/2017

TEDxBerkeley: Why it's Not Your Fault You're Sharing Too Much Online

A TED talk I was invited to give at TEDxBerkeley about the increasingly public digital lives we are being nudged into living. I discuss what's causing people to inadvertently reveal large parts of their lives online, why it's a problem, and what we can do about it.

2/6/2016

Harvard Club of Seattle: 125th Anniversary Dinner Keynote

A 1-on-1 discussion with Zillow CEO Spencer Rascoff about leadership and technology that I was invited to moderate for the Harvard Club of Seattle's Anniversary dinner.

6/26/2017

How to Scrape the Internet: Uncovering Privacy Leaks and Holding Tech Companies Accountable

A talk I gave at the Harvard Institute for Quantitative Social Science on what a privacy leak is, the dangers it can present to users, the root causes that cause it to exist, and how to pratically uncover leaks in the wild using a series of different tools and methods.

3/28/2015