Mark Labbe

This project had a small window of time for completion due to costs associated with keeping the acquired platform.

Working with a small team of developers and a database administrator, we extracted all of the data for the sites and transformed it to work within our existing platform. The primary roadblock here was the lack of access to documentation and original developers and the over-complicated nature of the platform's data. Much of the work was transforming content with in-line HTML, mapping section/category relationships to our own system, and ensuring that the final product design was a close match to the original.

The Java-based ad server had become bloated, and buggy and I was brought in, in part, to determine if it was worth fixing or if it should be replaced. Ultimately, looking at labor costs, the existing bug list, estimates from developers for fixes, and downtime statistics, I determined the system needed to be retired and that it could be replaced for far less money than fixing it and in far less time than fixing it would take.

Working with one other developer, I designed and built a new server which we launched geographically load-balanced on AWS EC2. The total development time was five weeks, which speaks to the relative simplicity of the server. We went from having regular downtime to virtually no downtime for the server. Eliminating original developers also reduced costs drastically. The stack was Node.js, Redis, and AWS RDS.

We followed that project up with a basic management interface using PHP.

The primary challenge here was data extraction because the search engine did not have a robust crawling function. The bulk of the work was developing code to extract data from web pages and databases and feeding it to the search engine. There were roughly 13 different platforms in play.

After inventing a line of guitar picks, I raised a small first round and with that money, set up the eCommerce site using WooCommerce and integrated it with multiple payment processors and a fulfillment service. I worked with designers and the marketing team to develop the design and established the production and stage environments as well as the RDS databases. I also set up the relationship with the manufacturer and launched the business within two months of spearheading the technology push.

I acted as a lead and managed a team of developers tasked with integrating the Endeca search into a large stock photography site. I worked closely with Endeca project managers and company producer to identify requirements. The project went through a rigorous testing process before release.

While StoryServer (at the time a relatively new product) offered a nice feature set, the editorial department was not thrilled with its data entry interface. Further, development wanted an easier way to build websites. So, I managed a team of developers who built a layer on top of StoryServer that allowed both developers and editorial staff to work together to build websites.

This was a web interface that allowed for content entry and organization as well as design changes to websites. The total development time was roughly two months. The site builds in this system were simplified by the ability to share content and design elements across sites.

The system was used to manage over 20 websites.

This API served as the back end for a campaign management system as well as for a data management platform. This system also had a built-in job queuing system and a notifications system.

The development was ongoing because of changing business needs. There were two primary developers, a QA engineer, and one product owner involved in the project.

One of the key requirements for the system was that code changes could be launched, specifically endpoints, without restarting the server. We also created an endpoint management system, partly, to help meet that requirement.

The existing website had been built on WordPress and had been built poorly. Further, given the need for a more dynamic site and development's skillset not aligning well with WordPress, we opted to create a React-based website.

The new system used Firebase as its primary data store and behind that was a MySQL database. For site management, we created another React interface. Data for the new site was extracted directly from the WordPress database as well as from spreadsheets that contained data on athletes.

I worked closely with a marketing agency to nail down the design and behavior. The time to first release was four months, and the site emphasized video content which included live streaming.