The Gregg & Mybrid Upgrade, Inc. (The GMU) was established in 2012 as an IT company comprised of the two founders, Gregg Yearwood (CEO) and Mybrid Wonderful (CTO). We have no employees but sub-contract out on a regular basis. Mybrid works remote from Maine exclusively and Gregg works remote when the work allows, but is otherwise on site in the San Fancisco, Bay Area.
Gregg Yearwood (CEO), B.A. Computer Science, Boston University, Boston MA, ‘92
After I graduated from Boston University in the 1990s I started my IT experience as one of the first employees at Geo Cities. Geo Cities was the first web site to provide personal web sites and was eventually bought by Yahoo!. Since Geo Cities I have focused primarily on QA and development lab IT and automation working for web and biotech companies in the areas listed below. I also work with Mybrid as a Python developer working on features and bug fixes.
Development environment servers and networks on the order of hundreds.
Development applications such as bug tracking and wiki pages.
Build engineering such as Atlassian JIRA and Jenkins.
Provisioning using VMWare, Kickstart, and Docker containers.
Hypervisor administration such as XEN, Virtual Box, and ESXi.
Python development system and integration tests using pytest.
Mybrid Wonderful (CTO), B.S. & M.S. Computer Science, UC Berkeley, Berkeley CA, '98
I started my IT experience in the early 1990s as a VAX system administrator. I went on to get two degrees computer science from UC Berkeley's College of Engineering with a concentration on relational databases. Since 2000 I have been both Senior Engineer and Project Lead for companies big and small. Below are listed some of the notable software projects representative of the work I've done over the years. In addition I work closely with Gregg as an IT administrator and provide backup support.
Charles Schwab: Login and authentication that was developed in C/C++.
Charles Schwab: First ever integrated sandbox for development of the Schwab web site. The software was developed in Perl. The project cobbled together Java, J2EE, C/C++ and 3rd Party application web sites that comprise the entire Schwab web site. This project allowed developers to fully test their work across the Schwab site in their sandbox.
Picso: Performance engineer integrating hundreds of image processing servers. The software was developed in Perl.
Affymetrix: DNA sequence haplotype translation table work in C/C++.
Affymetrix: JIRA/Bamboo administrator and build engineer.
Hitachi Data Systems: QA Automation framework for hundreds of development and QA servers. The software was was developed in Python 2.
Intrexon: Generic engineering of DNA application. The software was developed in Python 3.
Team Work @ Intrexon, 2018
Gregg and I spent four years at Intrexon working on a genetic engineering project building custom DNA for industrial products. The project required new sequence designs of up to thirty-thousand base pairs and forty sequence parts. The application had a proprietary algorithm that identified primer zones from experimental designs created by scientists. The DNA design software included the final sequence, the primers, and concentration levels. The application also output instructions for the first stage of robotics used in their LIMS system.
The software application required two developers to implement the entire software life cycle. Gregg and I collaborated full-time on this project where I was the developer and Gregg was QA automation along with bug fixes. We developed a novel test harness specific to DNA requiring three-thousand lines of Python code and relied heavily on the Python biopython package. Biology is messy and this resulted in a couple of gigabytes of regression data turned training data. We used bugs caught in production as both a regression data set and training data set for improved primer design. Testing was important because a single run of DNA construction in a high-throughput LIMS lab can easily cost fifty-thousand dollars involving thousands of new sequences. Bugs in the design code were expensive and reported meticulously to avoid any future cost of a failed run. After two years the regression testing was such that during the last two years Intrexon only saw a handful of minor bugs reported each year. This is the quality of software we can produce for you.
20,000 lines of Python.
LIMS CLI interface.
Scientific web interface written in Django for scientist use.
QA Automation Test Harness
Excel to Postgresql data management (ETL).
Biomax 3rd party software integration using ReST APIs developed in Django.