Tom Lovett

An entire REST API planned and built from scratch. The automated platform relieved the CEO/founder of the need for manual matching and scheduling, achieving a five-fold increase in customer base.

Peer Collective is a peer-to-peer counseling service that reduces the cost of mental health counseling by not mandating professional certification. Before this project, the CEO manually matched users with therapists, and clients-therapists were responsible for arranging meetings. I was hired to build an automated platform from scratch.

The CEO/founder was non-technical, so I architected all back-end aspects, including the database structure, API endpoints, and client-therapist matching algorithm. The custom-designed algorithm took in the user's needs and preferences for counseling and returned a list of counselors ranked by best fit. The app included integrations with Acuity Scheduling and Vonage, enabling users to book sessions with their counselors and the application to schedule video conferencing. A full test suite also supported the code, allowing later developers to quickly expand and modify the application.

A React/Node.js TypeScript app that enabled Upside employees to expedite applications for Paycheck Protection Program (PPP) Loan Forgiveness relating to COVID-19 loans. The user-facing app would walk users through the application process, while another front end served as an internal tool to manage these applications.

A strategic pivot from a travel company built during the height of the pandemic, myself and two client developers quickly spun up two separate front ends with a shared back end. The user-facing app was a single-page application (SPA) that prioritized speed and ease of navigation. The internal tool prioritized requesting, displaying, and manipulating rows of user data.

I wrote most of the customer-facing application, utilizing Material UI to build an intuitive SPA. The client developers and I paired to build the back end. The app was a success, allowing Upside Travel to keep employees on staff during a massive drop in business-related travel—their primary industry and source of revenue.

Replaced server-side rendered components with lazy-loaded React components, hitting GraphQL endpoints in a massively-complex codebase. The code was a success—the improved loading time encouraged our team to overhaul other components similarly.

EU regulations required Wayfair to display energy-consumption ratings for the thousands of its electricity-based products. The complex database lookup would slow down the render speed of the product page, so I was tasked with building lazy-loaded front-end components and their GraphQL endpoints.

I built the new front-end components behind a feature flag to not disrupt the existing server-side render components. These GraphQL endpoints were among the first of their kind at Wayfair. They needed to cover thousands of products, but my teammates nor I had prior experience with GraphQL, so I covered them extensively with tests and coordinated with the QA team for manual testing.

An API integration project with Checkr, a background-check processing service. I architected the API endpoints and database tables and built them alongside a client developer. I also redesigned the project to accommodate two strategic pivots. The project was a massive success, bringing in $60,000 in revenue in the first 90 days.

The client wanted to add a Checkr integration to their volunteer-management platform to allow users to request background checks. No one within the company had experience with 3rd-party integrations, and I quickly realized the project was unplanned and under-scoped. I offered to research Checkr and write a step-by-step build plan, which was approved, so I then wrote tickets for each step of the process.

A client developer and I built the first phase by working through the tickets. Users loved seeing the status of background checks, so Product wanted to allow users to request background checks. I replanned our API endpoints and database schema, wrote up another set of tickets, and built that functionality.

Then Product wanted providers to be able to upload their own background checks. I planned and built open-access API endpoints for the client, including new database tables.

A Ruby on Rails app to connect difficult-to-find people who stutter (PWS) with researchers who need participants for their studies. The app allows PWS to make themselves available to researchers while protecting their anonymity and only receiving invitations to studies in their proximity. For researchers, they have a pool of PWS who are otherwise nearly impossible to locate.

The app is SSR Ruby on Rails with ActionMailer and the Geocoding API integrations.

The client upgraded one of its Next.js applications, which used an obsolete design library, to a Remix application, which used the current in-house design library. As the sole contributor, I built front-end pages according to Figma designs, imported components from the client library (including troubleshooting language errors), and reported. I owned the project while the primary client developer was on extended leave and was praised for the progress achieved upon their return.