Hi! I'm Jason Davis-Cooke.

The start of my full time career came at Vistaprint in 2012. I joined a team with the primary responsibility of running the performance environment for the vistaprint.com release cycle. Every three weeks, as was the release timeline, I would perform load tests using HPs LoadRunner putting the most important workflows though the paces. In addition, I would manually capture code profiles of critical checkout workflows to ensure there were no bottlenecks introduced. In the weeks between performance testing, I was writing software to help me perform these performance duties more efficiently. The first of these tools was called VistaPerf. It was a C# application using WPF that was querying syslog data captured on the webservers for page loading times, and comparing the 7 day rolling average. If that average rose above a certain threshold, it would pop an alert for someone to look closer at that page. I caught a number of non-obvious peformance issues this way and allowed us to quickly remediate. Another application I wrote was also in WPF and was designed to ingest the code profiles I took and perform analysis on them as well as compare them to the historical data. It would show a heat map of the method calls in the code, and make any large spike in calls pop so I could flag them for further investigation.

The ICE Years - DevOps

Around 2014, Vistaprint reorganized into a parent company, Cimpress, and a portfolio of mass customization companies of which Vistaprint was one of them. I moved onto a team called Infrastructure Core Engineering, or ICE, whose primary responsibility was developing core infrastructure. At the time we were working on a migration away from the monolithic vistaprint.com application and into a more microservices architecture. To better facilitate the microservices world, this team was tasked with developing the services like CI/CD, monitoring and alerting, and logging.

• • Zanzibar
• • Uhaul
• • MCP ELK
• • Zamboni

API Management - Lead Software Engineer

Key Contributions

• • Cimpress Open Access Management
• • Cimpress Open Catalog API
• • Auth0 SME

I started at Indigo as a Principal Software Engineer on a team dedicated to building scalable tooling for the Data Science team running analyses on microbial samples collected on farms. On a team of 3 engineers, we built the Community Sequencing API, a typescript/graphql API with a React frontend that allowed for the upload of spreadsheets containing information about various studies done on fields to be stored in a postgres database for later analysis. We then developed a framework that allowed data scientists to configure and run analysis scripts in a scalable AWS environment. This started out as a docker skeleton project with instructions on where to put the analysis script so that it would run on container start, as well as conventions for installing job specific dependencies on container build. This allowed the data scientists to focus on writing the analysis code rather than needing to worry about how to build and deploy it, and also got them away from SSHing into EC2 containers to run these scripts manually.

A few months after starting, the VP of Engineering, who managed all 20+ engineers at the time, left the company. The CTO then decided that we needed to install an additional layer of management bewteen himself and the engineers. I stepped up to fill one of the gaps and began my management experience in September of 2018 with 2 direct reports at the time. Going into the last quarter of 2018, I was tasked with putting together a plan for hiring on my team going into 2019. Based on the roadmap I had created with our product manager and stakeholders and the available budget, I was able to open 3 positions to hire. I filled two of the three positions by the end of 2018. Early in 2019, one of the newly hired engineers came to me with an idea; AWS had released Step Functions which allowed for composable, scalable workflows. He believed this would be exactly what we needed to improve our existing pipelines for genomic analyses. I asked him to spend some time in the upcoming sprint to put together a proposal and if time allowed, a PoC for what this would look like. After coming back with a compelling argument, I greenlit the project and the Genomic Pipeline was born. This project ended up reducing the time it took for the data scientists to run the suite of analyses on their entire sequence catalog from 30 days to 7 days allowing for more rapid decision making for the microbial division of the company.

Also in 2019, Engineering was preparing for a rapid growth phase. The CTO wanted to take us from 30ish engineers to 100. I was a part of the management team that worked on creating the structures and processes that would allow for that type of growth including the creation of a well defined carreer ladder, a streamlined interviewing process, and onboarding plans. I was invlolved in many discussions on many different topics including office expansions, compensation plans, performance calibrations across the engineering team, and organization goals and metric tracking.

By the end of 2019, Engineering had grown rapidly and the Grain Marketplace vertical within engineering was attempting to reorganize in order to become more agile. This meant they were separating the front end development from the backend development, and a new Marketplace Platform team was created for the backend services needed to power the front end. These teams were mainly python developers and there were not enough of them to build all of the services that were needed at the time. I was asked if I could offer any assistance from my team who were still focused on the microbials work, and I ended up splitting that team into two teams, one of which would work on Marketplace Platform services, with my own time also being used to that end. Across the next year, my total number of reports grew to 12 across 4 different teams, and my teams gained a reputation for delivering. If there was a new service that needed creating, or a project that was stuck and needed help, I was assigned to it with one of my teams. We delivered a number of APIs across that time including an API for ingesting important documents needed for transacting on the Marketplace, an API for ingesting third party grain bid data for presentation in the Marketplace, and an API that allowed transportation carriers to define rates to transport grain within self defined geographical zones and then calculate the best rates for transporting from one zone to another. During this time I was responsible for the scrum processes for all of the teams as well as participating in a "scrum of scrums" to coordinate work across the engineering organization. I also performed weekly 1:1s with all 12 reports and handled their performance and career management.

Halfway through 2021, there was a project started to rebuild the Grain Marketplace from the ground up, and I was chosen to contribute to that team as a hands on engineer. I was tasked with developing the APIs needed to track user identity and define key business objects such as farms, grain elevators, and other entities needed to transact on the marketplace. We had a 60 day deadline to produce a working MVP to show to a potential partner and close a deal with them. As the only engineer working on these APIs, I had to make quick decisions about what the most important functionaliy was and were to cut scope in order to hit that deadline while still providing what was needed for the MVP. In the end, we successfully hit the deadline, presented the MVP, and closed the deal with the partner.

Coming out of the marketplace rebuild, I was then asked to lead and build the authorization system for the new marketplace. There was a strong desire to move away from the role based system that the original version of the marketplace used as it had caused many issues over the years. We decided to pivot to an attribute based authorization scheme and Open Policy Agent was selected as the technology to power that system. I had roughly one quarter to, on my own, build this authorization system from the ground up. The system had to allow other teams to define the policies governing access to their systems, provide fast authorization decisions when asked (the target I was given was 5ms), and have access to relevant domain data needed to make those authorization decisions. I delivered this system a couple of weeks into January 2022, just in time for user testing with our key partner. I spent the rest of my time at Indigo as the trusted voice and SME for the team responsible for this identity and access platform.

• • Community Sequencing API
• • Genomic Pipeline
• • Clio API
• • Transport Pricing Service
• • Marketplace Ingestion API
• • Identity API
• • Accounts API
• • Access Management API

I joined EQRx in 2022 with the goal of building out the Software Engineering organization from the ground up. At the same time, there was work that needed to be done immediately in support of the data science organization. I got hands on helping redefine a data model for an internal financial forecasting application that was in need of performance improvements. At the same time, there was a need for the development of an API for providing a critical application governing the distribution of drugs to clinical trial sites with the data needed to power the components of the app. I had some budget for hiring, but not a sufficient timeline to source and hire full time engineers. To solve this problem I turned to contracting firms, and was able to hire two nearshore contractors within a month. They came onboard quickly and immediately started ramping up on the project and were productively contributing shortly after.

Around the end of 2022, I began working with senior leadership on budgets and plans for growing the engineering organization according to the needs of our internal customers. This plan included a mix of more nearshore contractors and full time engineers. In addition, I worked with the Director of Data Engineering to design a unified data access layer which would provide the data scientists with a single entry point into all of the sources of data available to them including postgres databases and Google Big Query. This layer would provide all the necessary authentication and abstract away the differences in query languages between the different sources.

In an effort to consolidate varying documentation sources and make it easier for new engineers to onboard as well as simplify the data discovery process, I started a project to introdude Spotify's Backstage as a portal into EQRx's data platform. By early 2023, I had successfully deployed Backstage and demonstrated its value as a documentation hub. Shortly afterwards I began a campaign to migrate documentation from other places such as Sharepoint and Guru into Backstage with tagging to make it all easily discoverable from the documentation homepage.

Unfortunately, in May of 2023 EQRx announced a 60% companywide reduction in force which included the eliimination of my role as software became deprioritized and I never got to fully execute on my plans for organization expansion and development of the Data Access Layer.