Wellio Meal Planning Platform

Research, Writing, & Strategy

An AI-powered meal planning platform acquired by Kraft, I partnered with Wellio's team to help develop impactful yet practical solutions to digital meal planning. In addition to my neuroscience research and creative work, I've dabbled in AI, also holding a master's degree in Data Science. Despite not considering data science something I'm truly passionate about, this foray granted me invaluable insights into the inner workings of popular generative AIs like ChatGPT and DALL·E 2. It also expanded my technological imagination and polished my storytelling skills. I've become quite skilled at weaving together narratives from complex findings to convey a clear and captivating message, which was a large part of my work at Wellio.

At Wellio, I collaborated with the design, marketing, and engineering teams to improve the user experience of their recipe search engine using the power of AI. By applying image recognition techniques, we ensured users were presented with high-quality recipe images aligned with their search intent. I wrote about my work as research for my Data Science master's degree, telling the story of the business problem and presenting the solution strategy. I also outlined the implementation, analyzed results, developed conclusions, and directed Wellio's future product development.

Before Wellio, I developed an independent side project in a similar space, building an image-based nutrition calculator. Essentially, snap a photo of your meal and instantly get all the nutrition facts. Sounds like a simple idea, but the first step to such a complex problem is to recognize and categorize different kinds of food. So I built a rather amusing proof of concept: a pizza detector. Why pizza? Well, given its ubiquitous presence in the Western world, pizza served as the perfect test food for this endeavor. And if this idea rings a bell, you may recall a similar app featured on the show Silicon Valley. But rest assured, and please be sure to catch my sarcasm here, I had this absolutely genius idea first. While I personally developed the AI and machine learning (yes, I single-handedly wrote all the code), what I truly enjoyed was brainstorming the initial product concept, ideating on use cases, crafting witty writing, and assembling informative yet fun visuals for the project's website. Overall my aim was to create an entertaining and approachable narrative explaining a solution strategy for part of a larger, much more complicated problem that has yet to be solved. I currently host my code on GitHub and have gotten several requests from others to use my data and code for their projects.