Foodily (acquired by ask.FM)

While working on Foodily's Ingredient and Recipe Ontology, I conceived and implemented an exhaustive and scalable semantic map for the core product and maintained it throughout my time as a linguist, data editor and eventually as I became the company's first Product Manager. I worked closely with the engineering and marketing teams to develop a sophisticated search engine and content pipeline that was scalable, engaged users, and generated repeat traffic.

Foodily's homepage

Foodily's latest homepage combined the 2 search bar design and the magazine-style editorial content. Adding different editorial categories to homepage that changed daily proved to be a powerful way to increase engagement and to inspire people to search for more content without getting "search bar paralysis".

Original homepage

Foodily's original homepage featured only 2 search bars, one for users to to specify an ingredient, recipe, or diet, and another to exclude ingredients.

Curated Content

Recipe content aggregated from various websites and blogs across the web were featured on the homepage and on social media channels. The content was selected and curated by an editorial team that I hired and managed.

Semantic Search

Using NLP and various machine learning frameworks, my team stemmed, parsed, grouped, and classified ingredients and dish names in a way that made the millions of data points we indexed be more browsable and searchable by humans (specifically, foodies). The outcome was hundreds of dynamic filters and attributes that we used to plant the information architecture of the site.

Data-rich Content