Sifter: Building a personalized nutrition platform for safer grocery shopping

Client-Story-Sifter

Sifter is the brainchild of brothers Andrew and Thomas Parkinson – veterans of online grocery services. They imagined an online platform for consumers personalized to their dietary and nutritional needs, making it simple and safe to shop for groceries. 

Before Sifter, the Parkinson brothers were familiar with Egen from its work reimagining the cloud architecture for Peapod, the online grocery delivery company they co-founded. Delighted by the engineering and platform expertise that Egen provided, they returned to our experts to help build Sifter.

We understood the Sifter mission immediately. Sifter started with a consumer-focused business model, looking to build a personalized grocery platform to meet different consumers’ dietary needs and preferences based on factors such as their lifestyle, health, and allergies. 

We knew Sifter would need robust data pipelines to handle information on more than 25 million grocery stock keeping units (SKUs) to identify individual products from multiple partners. This data would be communicated over a seamless user interface that would recommend products based on preferences, diet, and lifestyle.

Egen helped the Parkinson brothers realize their vision. From building a complex machine-learning driven search tool to developing the user interface, we helped launch Sifter from the ground up.

To deliver personalized recommendations, we created a secure platform that exposes APIs available for consumption by third-parties as well as Sifter’s client itself. We automated the ETL pipelines that led to 8x faster catalog refresh and 90% reduction in live data indexing times. By implementing a serverless architecture, we designed a reliable, scalable platform for Sifter that saw a 99.98% uptime in one year, helping Sifter realize its vision of personalized groceries for all.

We have also built unique tools such as RecipeSifter, which uses machine learning to identify the ingredients in recipes and analyze whether those ingredients fit with a user’s dietary preferences. None of Sifter’s competitors have an offering that can match this.

Recently, Sifter is pivoting from focusing on serving consumers directly to working with other businesses. We are helping Sifter build capabilities and third-party integrations that it can sell to major food manufacturers, retailers, healthcare companies, and more. To date, more than 50 retailer data feeds have been added for Sifter with third parties, including brand names such as Walmart and Target as they continue to explore new possibilities.