About Me


Hello! I'm Dylan Estrada and I am a Software Engineer at Aurora Innovation. I enjoy working in all areas of the field and diving deep into challenging, complex problems.

In my free time I enjoy biking, running, playing Ultimate Frisbee, and hanging out with my friends.



Projects


These are some of the projects I have worked on in the past.
Raspberry Pi Image

Image Detection using Raspberry Pi

I utilized TensorFlow and OpenCV to create an Image Detection script that will send a "brake" signal to an LED if objects are detected in the area.

The Pi outputs its vision and calculated distance of the object to a web frontend app.

This project is one that I would love to continue in the near future, adding the Pi to a RC Car would be a very cool way to explore Autonomous Vehicles!

Fitbit Logo

RGB Control from Smart Watch

In this project I created a program to set RGB devices connected to a user's computer based on the user's heart rate data.

To get heart rate data I use fitbit's web api. RGB devices are detected and set using openrgb-python, an API created for the opensource project OpenRGB.

This project is still under construction. I'd love to add support for other devices like Apple Watch and Garmin.

Unix Image

Unix Operating System

I worked on a Unix Operating System in a Computer Systems class during the fall of my 3rd year.

My OS supported processes, threads, scheduling, a file system, virtual memory, and even multiple terminals!

This project was a great, if tough, way to get familiar with Unix systems.


Image of heartbeat

Smart Watch Heart Attack Predictor

In my IoT Devices class during the fall of my senior year I worked with a small team to simulate Smart Watches in AWS. Our watches would read a user's hearbeat, run a machine learning script, and then let the user know if they were at risk for a heart attack.

This project introduced me to AWS IoT, as well as taking me deeper into AWS Lambda and machine learning. AWS IoT really impressed me with its ability to simulate thousands of devices as if they were deployed in the field.

I'm looking forward to implementing it in my own projects, it might even be a good pairing with my RGB Smart Watch project!




Résume Timeline


This is my education and work experience timeline.

  Download My Résume

You can download my résume in PDF format if you would like.

   Download    View

  Programming Languages

Python
C++
C
SQL

x86 Assembly
System Verilog

  Libraries & Frameworks

PyTorch
OpenCV
Pandas

NumPy
MatPlotLib

  Tools

git
Github
Buildkite
Docker
Kubernetes
Kafka
PostgreSQL

Unix
Windows
AWS
GCP

  Relevant Coursework

Computer Vision
Internet of Things
Computer Security
Computer Systems
Algorithms and Models of Computation
Linear Algebra
Digital Signal Processing
Data Structures
Analog Signal Processing

  • Aurora Innovation

    Software Engineer II - Perception Metrics

    Pittsburgh, PA

    • Received company-wide award for quickly designing and implementing Perception’s large-scale Continuous V&V process leading to perfect pipeline health and 50% reduction in runtime from previous process
    • Created weekly job to update test datasets allowing for label fixes to be used 4 weeks faster while also providing a detailed report of how the changes impact metrics
    • Designed tool to track, visualize, and link to recent PRs which impacted Perception’s performance increasing change visibility and reducing regression triage time by over 90%
    • Created Buildkite pipeline with day-to-day and week-to-week comparisons of Perception performance with notifications of statistically significant regressed metrics sent to DRIs with a TP rate >.99
    • Curated new test dataset by coordinating with teams across Perception to narrow in on ODD, reducing dataset size by 30% while increasing signal in nominal and long-tail cases
    • Horizontally distributed key portions of the metrics pipeline, leading to a 300-minute speed-up of gating jobs and increasing developer velocity

  • Aurora Innovation

    Software Engineering Intern - Label Data Platform

    Pittsburgh, PA

    • Designed issueset structure and API which allowed users to create batches of label issues and submit them for relabeling, reducing turn-around time by 50%
    • Performed label service database maintenance, updates, and created documentation of team processes for future users

  • Caesar Research Group

    Undergraduate Researcher

    Champaign, IL

    • Assigned team members tasks based on experience levels, increasing team productivity and reducing turnover
    • Created low-latency, multi-user, Kafka based TCP server to allow for multi threaded communication between backend database, backend emulation, and frontend server
    • Redesigned APIs between Unity VR project and cloud network, reducing latency by over 30%
    • Compiled list of best practices to establish connection to backend cloud network for future onboarding and server expansion

  • Kohl's

    Software Engineering Intern

    Milwaukee, WI

    • Developed Jenkins Pipelines using Kubernetes, Ansible, and Docker to improve CI/CD operations and reduce runtime by 25%
    • Updated network to help internal customers easily view allocated network space reducing related tickets by 95%
    • Expanded firewall rules to allow for more consistent internal access while preventing unwanted connections
    • Demoed project functionality to internal customers and managers, including a recording to aid future users

  • Tri-the-Illini Triathlon

    Assistant Race Director

    Champaign, IL

    • Oversaw and delegated responsibilities to other members of the race organization committee
    • Coordinated with local police and governments about road closures and location of EMS teams
    • Created timetables for future race committees and streamlined process with local government

  • University of Illinois at Urbana-Champaign

    Student

    I graduated from theUniversity of Illinois at Urbana-Champaign, with a BS in Computer Engineering

    In Spring of 2020 I made Dean's List for the College of Engineering