Close

Michael R. Lane

Software Development Engineer

Download Resume

About Me

I'm a Software Engineer near Portland, OR. I have an MS in CS with a focus on AI and Machine Learning from Portland State University. I'm passionate about engineering elegant and performant solutions to big problems. I live for the opportunity to learn more and to perfect my craft. There are few joys greater than being an integral part of a successful team.

Experience

ChowNow

Sr. Backend Engineer

  • Architected a new microservice that allows ChowNow to integrate with Toast, a major Point of Sale provider for restaurants, using API Gateway, AWS Lambda, and DynamoDB
  • Used Python and Terraform to build out our new highly available, scalable, and inexpensive service on AWS which can serve 10,000 restaurants and 100,000+ diner orders per day
  • Mentored junior engineers on software best practices such as static and run time type checking, test automation, architecting for scale, and designing systems that are flexible and easy to change
  • Worked directly with stakeholders at Toast to demonstrate our integration for certification

Amazon

Software Development Engineer II

  • Used Typescript to build a new GraphQL service that allows specialized customer service agents to alert abuse prevention teams of abusive customers
  • Used Java to build out a highly available and scalable service that allows the customer service message bot to grant concessions to send customers to a specialized customer service agents for further verification based on machine learning predictions

App Annie

Sr. Backend Engineer

  • Used Python, Flask, and Postgres to build an machine learning-enabled service to provide users with relevant insights into the performance of their app on several metrics compared with their competitors
  • Used Pytest to create a new testing framework that increased our test coverage from 85% to 95% while reducing the lines of code by 15%
  • Architected a scalable and performant Unified Feed to replace App Annie's home page which combines data from many existing sources into a feed that prioritizes items based on our user preferences and activity
  • Architected a system whereby new internal data sources as well as data from future external sources (from partners, news sources, blogs, etc) can be quickly added into the Unified Feed system on the home page

Amazon

Software Development Engineer II

  • Architected a global, highly available, machine learning-enabled web service that supports all internal and outsourced Amazon Customer Service Agents and their managers world-wide
  • Used Typescript, React, AWS AppSync, AWS Lambda, and DynamoDB to build highly available, performant, and secure web applications for Customer Service Agents
  • Used AWS Cloud Development Kit and Typescript to define our infrastructure as code and leveraged the AWS CodeSuite to build fully automated CI/CD pipelines
  • Designed and built an Automated Acceptance Testing service using Typescript, Jest, and Cucumber
  • Architected and a highly available click stream logging and analytics solution using AWS Kinesis Firehose, Kinesis Data Analytics, and AWS Managed Elasticsearch
  • Spearheaded my team’s effort to push Amazon into the future by making legacy Java services securely available to next generation AWS-based clients
  • Used custom extensions to Java Spring to maintain legacy enterprise services

Perch Security

Software Engineer

  • Use Python, Django, Postgres, Redis, and Docker to build and extend the Perch Security web service
  • Using Elastic Search, Logstash, and Kibana to architect a data lake and machine learning analysis platform for petabytes of security indicator data
  • Use Javascript to create custom modifications to Kibana that is tailored to the needs of security professionals who use Perch

Uncorked Studios

Sr. Python Developer

  • Using Python, Django, MySQL, and GraphQL to rebuild a web service from the ground up to allow the client to scale by 10X
  • Building serverless ETL user clickstream pipeline in Django using AWS Kinesis, Firehose, Glue, Lambda, Athena, and S3 to capture rich user activity data for future Business Analytics

Nextas America, Inc. (Startup failed to acquire funding)

Software Engineer

  • Using Python, Keras, GPU accelerated Tensorflow, and C++ to R\&D novel deep learning approaches that replace legacy computer vision tasks with algorithms that are faster, more robust, and far more accurate
  • Built a novel deep neural network using Python and Keras to generate a depth map from a pair of stereo images

Portland State University

Graduate Teaching Assistant

  • Instructed students on machine learning topics such as neural networks, bayesian networks, reinforcement learning, support vector machines, ensemble learning, unsupervised learning, and classifier evaluation
  • Provided instructions and guidance to students implementing a term-long project to build a user-level threading library using C and Assembly

CDK Global

Software Engineer Intern

  • Developed REST API service using Python, Flask, and Cassandra that supports 40,000 phones making hundreds of thousands of simultaneous connections
  • Built system using Machine Learning to allow 3 Raspberry Pi receivers to track the locations of 6 BLE beacons to an accuracy of within 6 meters

Acquia

Support Engineer Intern

  • Fixed bugs and improved functionality of a Remote Administration script using PHP, Bash, and Drupal Shell (Drush)
  • Created an automated task scheduler using PHP, SQLite, and Drush that reduced time required for manual tasks by 80% and reduced the time spent running a weekly mass update task from 3 hours to 5 minutes

Portland State University

Technical Course Support Specialist

  • Supervised 2 to 3 instructors and 20 to 30 students during CS course labs teaching programming basics, algorithms, data structures, and OOP principles using C++ and Java

SmugMug

UI/UX Designer and Developer

  • Customized SmugMug sites using for important customers and partners
  • Programmed help pages using PHP, HTML and CSS to document and explain site updates for end users

US Air Force

Air Mobility Liaison Officer

  • Coordinated major US Air Force airlift operations in support of US Army operations for peacetime and wartime missions
  • lanned and led wartime replacement operations of one 3,000+ soldier brigade with another in Mosul, Iraq
  • Developed a plan in Iraq to establish flight operations at a remote air base resulting in a reduction the need for deadly convoys

US Air Force

KC-135 Evaluator Navigator

  • Planned and executed high priority air-to-air refueling missions worldwide
  • Instructed and evaluated KC-135 navigators to ensure they could complete missions safely and effectively
  • Conducted nearly 100 air refueling missions over Afghanistan between October 2001 to March 2003
  • Executed over 100 air refueling missions over Iraq between March 2003 and May 2003

Education

Portland State University

June 2017

Master of Science in Computer Science, emphasis AI/Machine Learning

University of Phoenix

Master of Science in Computer information Systems

Illinois Institute of Technology

Bachelor of Science in Mathematics

Projects

Face Detection & Recognition

Face detection and recognition has now reached the status of a classical problem in Computer Vision. There are many different ways of recognizing and then classifying digital images of faces. This project implements the Haar Face Cascade algorithm for face detection. For face recognition OpenFace, seven different classifiers from Scikit-Learn (all of which were trained on the LFW data set), and Amazon’s AWS Rekognition were implemented.

View Project

GeneraList

GeneraList lets you use Amazon's Alexa to create and use lists that hold any type of data. Ask GeneraList to store grocery items as you think of them during the week and then access those items when you're on the go at the supermarket. Or give generalist a list of instructions that Alexa can read off to you later as you ask her to. Have a recipe that you love? Let GeneraList read the instructions to you as you're cooking.

View Project

Classify UCI Spambase Data using SVM

In this class project, I used Python, Pandas, and Scikit-Learn to classify the University of California Irvine Spambase Data Set. The data set contains 4601 rows with 57 columns of attributes and 1 column that contains the classification, 1 for spam and 0 for not spam. The experiments were to run a SVM on the data set and then to experiment with feature selection.

View Project

Skills

Get in Touch