It can take several years to develop new consumer products and packaged food. Companies must research, design and prototype the product. People expect trends to land faster in a socially connected society. Founded in 2018, Ai Palette uses machine learning to help companies spot trends in real time and get them retail-ready, often within a few months. Today, the startup that has clients such as Danone, Kellogg’s and Dole announced it had raised a $4.4 million Series A led by Exfinity Venture Partners and pi Ventures. They will both join Ai Palette’s board.
Anthill Ventures was also part of the round, as well as Decacorn Capital and AgFunder (returning backers in food technology ventures) It brings Ai Palette’s total raised to $5.5 million, including a seed round announced in 2019.
Ai Palette has its headquarters in Singapore and an engineering center in Bangalore. The company’s initial customer base was in Southeast Asia. It then expanded into China, Japan and Europe.
The company’s initial customers were in Southeast Asia and India. It then expanded into China, Japan and Europe. Ai Palette is capable of predicting trends in consumer packaged goods (CPG), and supports fifteen languages. The company plans to use its funding for expansion into new markets as well as filling engineering and data science positions.
Ai Palette was founded in 2018 by chief executive officer Somsubhra GanChoudhuri and chief technology officer Himanshu Upreti, who met through Entrepreneur First, the “talent investor” that recruits and teams up potential founders.
GanChoudhuri was a sales representative and marketer at Givaudan (the world’s biggest manufacturer of flavor and fragrances), before joining Ai Palette. He was able to observe how innovation happens for many different types of consumer goods, from packaged products to fast foods and snacks. He saw that two years of product innovation cycles were not enough to meet the demand for many companies. Upreti is an expert in big data analytics and machine learning. He has previously worked for Visa where he created models that could handle petabytes.
Foresight Engine is Ai Palette’s first product. It tracks trend information like flavors or ingredients and analyzes their popularity to predict how long they will be in demand. The engine also finds “white spaces opportunities,” situations in which there’s not enough demand. GanChoudhuri, for example, said that the COVID-19 epidemic has altered the way people eat. They now eat health snacks as many as six times per day in front screens. This gives companies the opportunity to create new products.
Upreti said that Foresight Engine provides contextual information. For example, is the food eaten at home or in a café? Are you consuming the product individually or in a group setting? What is the most popular at children’s birthday parties? Images provide details about product formats and pairings for a particular product, or ingredient.
Data sources include social media, search engines, blogs, recipe, menus, and company data. GanChoudhuri said that data sets most relevant to each market get prioritized. For example, a local recipe and a delivery service app. They are monitored over time to establish a growth trajectory with high levels of confidence.
Ai Palette’s technology has been used to create new products. For example, brands want to introduce a brand new potato chip flavor or soda in a particular country. The Foresight Engine can help them see which trends are on the rise and determine which products have the potential for being long-term favourites so that they do not invest in products that may lose their popularity quickly.
Ai Palette clients used the tool to respond to changes in consumer behaviour patterns and trends during the COVID-19 epidemic. It is not surprising that many people are looking for healthy foods or those which boost immunity. In the United States, acerola, yerba matte and lemon are in high demand.
GanChoudhuri said that taste, in China, is more important than health. People are seeking food that restores a sense normalcy. As people in India continue to deal with the pandemic, they are looking for longer-lasting products. However, many Indians also want interesting snacks to relieve the boredom caused by lockdown. Kimchi and other Korean flavours are becoming increasingly popular.
Ai Palette is able to handle many languages, which makes it stand out from other trend-prediction systems that use machine learning. The platform currently supports English and simplified Mandarin as well as Thai and Vietnamese. It also has Bahasa Indonesian and Bahasa Melayu languages. Plans to expand the support for Tagalog and Spanish in future, especially when it targets Latin America, Mexico and Europe.
Publiated at Tue 24 August 2021 22.37:13 (+0000).