A recap of some of my favorite projects
Query Builder was a project for interacting with the Best Buy API documentation. We built this after feeling the frustration of developers at hackathons trying to work with the API quickly, but having to always refer back to the documentation. They say necessity is the mother of invention. Hackathon developers were slowed in writing valid API requests by having to read lengthy documentation. This tool is interactive and faster to use for simple requests.
Best Buy on IFTTT
The IFTTT interface lives at the intersection of fun and functional, allowing users to create custom notifications about the events that matter to them. Want to know when a price changes or your favorite product is available? Interested in popular products in a particular category? IFTTT will drop you a note ass soon as the info is available. We built the Best Buy channel on IFTTT as an early step in connecting Best Buy to the IoT space, crafting customized alerts for changes in product status. For example, when your refrigerator senses that a water filter needs replacing, you can arrange automatic shipping and avoid the hassle of tracking down parts. Or on the fun side, you can change the color of your Phillips Hue light to Best Buy blue when there are new popular products.
Best Buy API Developer Program Transformation
I developed the vision, product strategy, roadmap for relaunching the Best Buy API program, essentially turning an afterthought into an essential piece of retail operations. I pitched to the executive team for funding to re-build the product, engineering, marketing and operations teams in a more efficient way. By replacing an outsourced team of 100+ with an internal team of 23 and instituting a zero-bug backlog, we saved the company $7.5M and launced 8 new products, including a rewrite of the developer.bestbuy.com website. API key enrollment increased 44% and test-first practices paved the way for the first automated continuous delivery pipeline in the company.
As a 54-hour hackathon project at Startup Weekend, my team tried to unite mentors and mentees in the moments that they need support by building a mobile app to close the gender gap through micro mentorship. See the pitch here. See our hustle on Twitter here. #SWTC6
I loved this project because we took a truly experimental and pragmatic approach. We made available a small sample of what seemed to be the most useful of the Best Buy recommendations as a public API. Although we had several production customers, there was little traction after launch beyond those core users, so we devoted our time and money to other experiments. This allowed more exciting products to come to life while keeping the Recommendations APIs lean and simple.
Best Buy eBay Store
This project was focused on creating a Best Buy store on eBay to expand our consumer market. Launched in 2014, we added 27k SKUs and quickly hit $20M in revenue. This platform allowed us to upload the product listings and product details using our Product Catalog APIs and manage product deprecation and sales with our Commerce APIs. The future of the platform was intended to be used with a variety of marketplaces like eBay and Amazon.
Best Buy dotcom Platform Transformation
The early transformation architects engaged me as a project manager fairly early in the effort. I focused on socializing the process and technical goals by establishing transparent relationships across the organization. I was a big champion of showcases (demos) by my team (as mentioned below under the Magellan project) and instilling agile and XP practices. Attendance at our showcases increased by 1000%, and we gave a two-week checkpoint with the business teams, both were was critical in selling these ideas and led to some indisputable improvements. Outsourcing became insourcing, and quarterly releases that formerly lasted 24+ hours became weekly lasting a couple of hours. I had the opportunity to work on transforming the product catalog APIs, product description pages (PDPs) and product listing pages (PLP).
Fun. Magellan was fun. As the project manager, I blazed through old crumbling road blocks left by the previous outsourced model to allow the developers to focus on “freeing the product catalog.” I built relationships and used my bulldog face (only for good, never for evil!) to move mountains. My eight-person team (all engineers, with one product manager and one project manager) used many XP and Lean practices, including test-first and pair programming, to provide the catalog data to all of Best Buy in a robust and highly scalable manner. We maintained excellent code quality, endured few defects and tolerated no bugs in our backlog. We transformed the Best Buy product catalog from a single source DB to a key-value data store (allowing zero-downtime deployments, fail-over and redundancy) that could serve the entire company rather than only the website. We only had one QA person in five years (who left after just a few weeks, only working 25% time because of the test coverage).
Boston Scientific Latitude CRM
I've never learned so much about medical devices and Japanese culture. I was the system analyst and was responsible for managing all of the system requirements of the CRM for heart patients with implanted devices and their doctors. The business owners and customers were in Japan and the engineering team was in the US. We built the system that allowed patients to check in with a device at home, through which doctors are able to monitor the results, reducing the trips to the doctor's office for the patient and improving data for the doctors.
Best Buy Mobile Phone Upgrade Checker
As a small experiment in 2008, we hacked together a quick prototype in three weeks that would allow Best Buy store employees at Best Buy to check phone upgrade eligibility for customers over the busy holiday season. If customers weren't eligible yet, they could opt in to get text message reminders as their eligibility approached. For the prototype we screen-scraped mobile carrier sites with data inputs from customers and then integrated with MyThum to manage opt-ins. The test had incredible results: we helped customers better understand their eligibility dates and created leads for selling phone upgrades. The project grew into building an open API to use across platforms (store, dotcom, etc.) and connecting to the carriers by API for more reliable data.
In 2009, I worked with Dashwire, a Seattle startup, on building a mobile-to-web connected services platform called mIQ. The mIQ platform allowed users to use their phones from their computers, back up phone data, and easily switch devices without losing their data. The second release was built for Best Buy, which was an early investor, and I managed this project. Dashwire was acquired by HTC in 2011.