/resume

My Resume and Stuff

Ashutosh Priyadarshy

Education

University of Virginia 2012
B.S. Electrical Engineering with High Honors
Concentrations: Signal & Image Processing and Control systems
GPA: 3.72/4.00

Work Experience

Sunsama

Co-Founder & CEO - San Mateo, CA & Portland, OR
October 2013 to November 2015

  • Technologies used: Python, NLTK, Flask, Javascript, Meteor, AWS, Git.
  • Accomplishments: built product, built team of 5 engineers, raised money, released product, acquired thousands of users.
  • E-mail AI: Used Python/Flask/NLTK to build an e-mail AI that automatically handled inbound requests like "find a time for Joe and I to meet next week".
  • Scheduling Engine: Built world's best scheduling engine for recommending available time. Accounts for arbitrary number of users, preferences, calendars, location and travel times. No other calendar/scheduling product has a scheduling engine has sophisticated as Sunsama's.
  • Date Range Parser: Built a custom Python library to convert text string like "between next tuesday and next Friday morning" into arrays of datetimes. Used NLP, regular expressions, and NLTK's grammar trees.
  • Chrome Extension: Built a JS chrome extension that added a command bar to the GMail DOM. Allowed users to automatically schedule meetings with their email contacts.
  • Contact Recommendation Engine: Built recommendation engine in python to suggest likely contacts based on contacts, mutual events, shared contacts, shared calendar and our internal user graph.
  • AngelList scraper: built AngelList scraper to automate outreach to angel investors.
  • Lead-enrichment Project: Scraped Slack's subdomains to get list of companies that matched a tech-stack profile we were looking for. Current products e.g. Datanyze aren't able to do this.
  • Slack Bot: Built a slack integration and slack bot to perform automated outreach to user's in public slack communities.
LUMOBody Tech

DSP and User Interaction Engineer - Palo Alto, Ca
October 2011 to July 2013

  • Technologies used: Matlab, ANSI-C, Python, Signal Processing, BLE, PIC Microcontroller.
  • Accomplishments: Built best-in-class pedometry algorithm that was used in clinical research and recruited and hired first two engineers.
  • Pattern Recognition Module: Wrote DSP algorithms for gait analysis and activity recognition. (Pedometry, calories, distance, impact) for the product's onboard microcontroller. Algorithms were prototyped in Matlab/Python and ported to C when ready for production. I also ported the functionality of Hidden Markov Model based prototypes to a DSP based approach to reduce computational load.
  • Height Detection Algorithm: Wrote adapative (machine learning) algorithm to estimate user height from accelerometer signals.
  • Context Adaptive Algorithms: Built models to re-parameterize pattern recognition module's parameters based on a change in user's context e.g. location, desk, etc.
  • Sedentary Behavior Research: Coordinated clinical research with Stanford university to test algorithms I developed.
  • Hardware/Sensor Testing: Built custom testbeds to ensure hardware robustness and measure various signal outputs. Performed tests in office and on the production line. Significant experience in testing and using BLE based devices.
  • Shopify/Volusion Backends: Built Flask/Heroku backend's to perform thousands of fulfillment offers.
  • SDK Examples: Used the Lumoback SDK to build sample iOS apps.
EEMe labs

Founder - Mountain View, Ca
June 2012 to October 2013

  • Technologies used: Python, Objective-C, NLTK, Cocoa, iBeacons, Location services, Twilio, PostgreSQL, SQLAlchemy, CoreBluetooth, iBeacons.
  • iOS App: Reachd: reachd used Natural Language Processing to generate automatic,intelligent and context aware SMS responses to questions about a friend's location, ETA, etc. Built backend services and NLP logic with Python/Flask. (Previously) available in in App Store.
  • iOS App: WalkBack: Created WalkBack, a product for student safety while in college. (Previously) available in App store. Won SXSW's Student Startup Madness with this.
  • iOS App: YOGG: Yogg was an experimental app that allowed you to use Siri-type location reminders for people you worked with. It understood requests like "Have Joe call me when reaches SFO." and would automatically cally both parties when Joe reached SFO.
  • APBeacons: Built general purpose iOS libraries that mimiced all of the functionality of the iOS's built in iBeacons library but allowed transmission and discovery while apps were backgrounded which was not allowed in iBeacons. I basically re-wrote iBeacons using only CoreBluetooth. This project is ongoing.
  • Sales Automation Tools: Built e-mail automation tools for cold e-mailing leads similar to SalesLoft Cadence and Outreach.io in Python.
Virginia Image and Video Analysis

Student Researcher at the University of Virginia
September 2011 to May 2012

  • Technologies Used: Matlab, Differential InSAR.
  • Developed MATLAB simulator for Differential InSAR to model sinkholes.
  • Image Segmentation and clustering of Highly Confluent Bovine Endothileal cells.
  • Taught iOS development to lab group.
United States Naval Research Laboratory

Intern in Tactical Electronic Warfare Division - Washington, D.C.
June 2010 to January 2011

  • Technologies Used: Matlab, C++, CMake, Visual Studio.
  • Porting MATLAB to C++: Re-implemented most of Matlab's Digital filtering functions and filter design tools into highly performant C++ code.
  • Radar/Missile Simulator: Developed code for a radar and missile simulator built in modern C++ with Boost.
  • Simulation & Data Backend: Built HDF5 data store for simulation data.
Azure Summit Technology

Intern in Electronic Warfare - Fairfax, Va
June 2011 to August 2011

  • Technologies Used: Multirate Signal Processing, Matlab, Octave, Scipy/Numpy, Python, Time-Frequency Analysis, Spectrograms, Filter Banks, Digital Signal Processing, Fixed Point Analysis, Cognitive Radios, Linux.
  • Cognitive Radio Adaptive Polyphase Channelizer: Built Matlab simulations of Azure's Cognitive-Radio, Electronic Warfare adaptive polyphase-channelizer used for interference cancellation, highly efficient signal reconstruction and anti-jam tactics. Performed real to I/Q conversion in a computationally efficient manner to simulate channel equalization.
  • Fixed Point Filter Design: Simulated the effects of fixed-point arithmetic to observe implementation losses in FPGA systems using Matlab/Octave's Fixed Point toolbox.
Aster Data Systems

Intern in Product Development Team June 2009 – August 2009

  • Technoligies Used: MapReduce, SQL, Regular Expressions, Java, Massively Parallel Databases, Linux.
  • nCluster Simulations: I wrote complex PostgreSQL queries to perform the sequential, time-based, pattern matching analysis that Aster's nCluster could perform in just a few lines in order to demonstrate the value of nCluster and profile it against native PostgreSQL.

Skills and Knowledge

Languages & Tools in Production Environments

  • Python: [NLTK/Flask/Numpy/Scipy]
  • Objective-C [iOS/Cocoa]
  • Javascript/Node [Meteor]
  • MATLAB/Octave [DSP/Image Processing]
  • MongoDB
  • RegEx
  • C [Microcontroller]
  • SQL [Postgres/MySQL/Teradata-Aster]
  • C++ [Boost Libraries]
  • Git & Subversion
  • XCode, VIM, Eclipse
  • Heroku
  • AWS [ElasticBeanStalk/S3/EC2]

Language & Tools for Fun:

  • Haskell, Julia, R
  • Arduino
  • Weka
  • Natural Language Processing

Relevant Course Work

Programming/Software:

  • DSP Laboratory
  • Software Development Methods
  • Neural Network Models
  • Computer Architecture

Image Processing:

  • Digital Signal Processing
  • Digital Image Processing
  • Nonlinear Control Systems
  • Digital Control Systems
  • Linear Control Systems
  • Neural Network Models
  • Medical Imaging Modalities

Mathematics:

  • Bifurcation, Chaos and Stability in Nonlinear Dynamical Systems
  • Nonlinear Systems
  • Linear Algebra
  • Discrete Mathematics I
  • Differential Equations

Products Built

  • Sunsama: Founder, led team of 5 engineers. Built original versions and various MVPs including an e-mail based assistant that auto-scheduled events.
  • WalkBack: Mobile-app for student safety. App Store
  • reachd: Context aware and auto-response SMS app. Link
  • LUMOback: Second full-time engineer. Wrote DSP and Machine Learning algorithms for activity recognition and health tracking.

Open Source Projects

  • APBeacons: iOS 7 provides developers access to iBeacons. When using an iOS device as a beacon you are unable to broadcast with the app being in the background. APBeacons was designed as a workaround to this issue. APBeacons uses CoreBluetooth in both the peripheral and central mode simultaneously in order to make background broadcasting possible.
  • GoogleCalendarV3 is a python wrapper on Google Calendar's API. This was downloaded 700 times in the past month.
  • email-att is a Meteor package that let's you send e-mail with attachments, this was originally not supported in Meteor's core implementation. It currently has 504 installs.
  • APKeyboardCategories is re Objective-C categories that add useful and common functionality that almost any text input field in an iOS app would need.
  • APResponseParser a simple utility that allows a user to retrieve subdocuments from JSON that is arbitrarily deep without having to employ checks for Nil values at every step. This is one of the first pieces of iOS code I wrote so I asked for help improving it via public code reviews: Stack Exchange Code Review
  • AnisotropicDiffusion is just a small Matlab script that performs Anisotropic Diffusion on an image. Examples show how this filtering method can clean up images of Neurons.

Publications

Teaching Experience

  • ECE 3209 – Electromagnetic Fields: Taught four full lectures, held weekly recitation.
  • BME 4641 - Bioelectricity: Undergraduate Teaching Fellow, held weekly office hours.

Competitions

  • SXSW Student Startup Madness – First Place for WalkBack
  • Darden Business Plan Competition – First Place for WalkBack

Awards

  • Chairperson’s Award ECE – awarded to outstanding students in the ECE department.
  • CIT Gap Funds “GAP 50 Finalist” - honors both the emerging and repeat entrepreneurs who will develop the next generation of high-growth companies in Virginia.
  • Rodman Scholar - The Rodman Scholars Program is the honors program for the top 5-6% of each class.
  • Jefferson Public Citizen - An academic honor and grant for students to perform community based research.

Older Projects

These are projects or research I performed during college and highschool.

CalmingTech Laboratory

Student Researcher
July 2011 - August 2011

  • Helped implement an algorithm to extract breath rate from accelerometer data. This product is now commercialized as Spire
Biogas Systems

Jefferson Public Citizen - Bluefields, Nicaragua
January 2010 - August 2010

  • Gathered data and information about the use, maintenance and issues involving biogas-systems in Bluefields, Nicaragua in order to create a market for the sustainable technology.
  • Used a human-centered approach to prototype solutions that will aid in the increased use of biodigesters.
Bioinformatics Project

Undergraduate Research Project - University of Virginia October 2009 - May 2010

  • Implementation of mass-spectrometry analysis algorithm in the statistical computing language R for biomarker discovery.
Neural Engineering Laboratory

Student Researcher at George Mason University September 2007 to May 2008

  • Performed electrochemical Studies of highly sensitive electrodes that interface with a rat’s brain.
Galileo Sequences

Summer student at Lynchburg College July 2007 to August 2008

  • Developed expansion of Classical Galileo Sequence Problem.
  • We developed a formal proof for sequences similar to Galileo's with variable arithmetic gaps and also when the terms are raised to any power k.
  • We were able to derive a general expression of the infinite expansion of binomials that expresses the value to which the sequences converges to based on the power that each term is raised to.
  • Problem Formulation published Math Horizons.
General Dynamics: Advanced Information Systems

Student Participant September 2007 - May 2008

  • Developed software for a Sonar Tactical Simulator working in a small team of four students. We made extensive use of UML diagrams and the Model-View-Controller design pattern before implementing the simulator in the Java language.

Contact Information

Analog:

2315 NE Couch St. Portland, OR 97232 (703)-725-9158
skype: ashutosh.priyadarshy

Digital: root@ashuto.sh | LinkedIn | Twitter | GitHub | Facebook

Fast Navigation:

Education
Work & Research Experience
Skills & Knowledge
Relevant Courses
Products Built Open Source Projects Publications
Teaching Experience
Competitions
Awards
Older Stuff
Contact