The Economic Times daily newspaper is available online now.

    Can carrot be paired with onion extract? A machine has the answer

    Synopsis

    Spoonshot uses food science, machine learning and a repository of vast data on food to predict food combinations for FMCG companies. By combining food science and machine learning to predict flavor combination trends, it aims to give companies an early-stage innovation product.

    iStock-475459097FINALGetty Images
    Kishan Vasani said that Spoonshot is starting with a focus on North America, but by 2020 end, it aims to launch in other markets with a preference for China.
    Hot Startups 2020
    At the turn of the decade, Kishan Vasani joined Just Eat Co in London as the Head of International Marketing. It was during his stint in Just Eat that he realised that his food experiences didn’t always match the restaurant ratings and reviews. “For example, one of my favorite dishes served by a restaurant in Bangalore has a restaurant rating of 3.1. Conversely, I have been to several highly rated restaurants (4.0+) and thought the food was average at best,” the 38-year-old entrepreneur told ET.

    Vasani thought maybe individual dish ratings might solve the problem. However, he came to the conclusion that taste is personal and the dish or restaurant ratings don’t make sense unless they were made by someone who has the same taste as the reader. A more innovative approach was required here to understand customer tastes and preferences and therefore provide better insights.

    Vasani made his exit from Just Eat in 2015 and later in the same year formed Spoonshot, which was earlier known as Dishq, after connecting with Sai Sreenivas Kodur through LinkedIn.

    With emerging technologies such as machine learning (ML) and artificial intelligence (AI) taking over almost every sector, it made no sense to leave the food industry behind. The idea for the Bengaluru-based startup was to help FMCG companies become the trendsetter by providing inspiration for new products.
    Founders--Sai-(Left)-+Kisha
    (Founders: Sai Sreenivas Kodur and Kishan Vasani)

    The wonders of food knowledge
    It has never been easy to track customers’ tastes and needs. There is always too much data and too many options. Research and insight providers often rely on lagging indicators and archaic methodology offering stale ideas which they further sell it to everyone, thereby destroying the competitive edge. However, Spoonshot’s insights are derived from its vast knowledge repository and analyses billions of diverse data points from over 900 sources in the domain of food which include scientific research, food news portals, menus, recipes, food communities, data related to flavour, sensory, ingredients and others. Adding a pinch of novelty, the firm has named the feature #foodbrain

    “With these knowledge repositories, our goal is to ultimately replicate human cognition in the domain of food,” Vasani said.

    The platform has three major services- ingredient networks, startup explorer, and trend trek.

    Ingredients Network enables you to search the ingredients for recommended flavour combinations. For example, if you type ‘carrot’ in the Ingredients Network search options, getting results such as apple or bananas would be useless since people already know about these combinations. Spoonshot’s job is to bring you flavours you haven’t heard or tasted, but makes sense because of food chemistry. Therefore, the search will result in unusual combinations such as mesquite powder, onion extract or saffron.

    You can further explore the combinations and see, on the basis of scores and percentages, parameters behind the pairings such as novelty, flavour, trend, popularity, and health. In simple words, if you have a higher score in novelty, it means the combination is pretty unique and not known to many. Similarly, a higher health score indicates that the combination will be very nutritious.

    Here, the Startup Explorer feature will go through ‘banana’ and let you browse through its database to show the combinations and pairings are already in use.

    1 (1)FINAL
    Food chemistry: If you type ‘carrot’ in Spoonshot’s ‘Ingredients Network’ search option, it will give you unlikely food pairings such as onion extract, bay leaf, and mesquite powder which may not sound delicious or healthy but makes sense because of food chemistry.

    Meanwhile, the Trend Trek feature will serve your wanderlust wishes and allows you to take a tour of the world see what flavours combinations and food innovations are popular in which country.

    “We have designed the platform so that users do not have to ask it questions, but are continually given answers or insights that are personalized to the user's needs,” Vasani said.

    The technology used here - Computational gastronomy and nutrition science. Computational gastronomy refers to the study of ingredients to understand how and why they are used with each other.

    “We extend this by collecting data related to flavor, sensory and texture. Nutrition science refers to break up of nutrition for each food and diet. We also enrich this with data on how a style of cooking influences nutritional value,” he quipped.

    He added that using this data is how they prepare the Ingredient Network feature which defines how the ingredients are related to each other in terms of taste, smell, nutrition, substitution, texture, etc.

    “This serves as a fixed knowledge base that acts as the backbone to build our applications,” he said.

    Apart from this, the firm also has a trend datastore which tracks every food data along with two key factors- time and geo-location, which helps them understand which ingredients are trending in which location.

    Disrupting the food-tech industry
    Despite working in the food tech sector, FMCG was a completely new space for Vasani and Kodur. To understand it better, the duo spent nine months exploring the sector and understanding problems and opportunities that come with it. “We met with over 130 industry professionals as a part of this exercise. This meant that our product development was validated and when we presented early solutions, customers were bought in and excited by our tech,” he said.

    Untitled3
    A list of FMCG products made through the unlikely combinations recommended by Spoonshot.

    While the first year was bootstrapped, the startup has, till date, raised nearly $700,000 in funding from High Net-worth Individuals (HNI) and investors.

    Brett Brohl, MD, Techstars Farm to Fork Accelerator, which has invested in the firm told ET, “Spoonshot enables big consumer brand companies to proactively understand future trends. Companies that take advantage of these insights will have a distinct competitive advantage for new product innovation, R&D, and even early stage M&A activity. Spoonshot has the ability to become a must-have insights tool for every consumer packaged goods company. Further, I feel the co-founders have the experience, drive and ability to build a great company.”

    The firm launched its product on October 1 this year, prior to which they were in beta, which over 40 companies had already signed up to participate. Vasani said that Spoonshot is starting with a focus on North America, but by 2020 end, it aims to launch in other markets with a preference for China.

    “We believe our technology (#foodbrain) and the platform will become the de facto choice for the food industry, across FMCG and restaurant/food services, for being on-trend and providing customers with the choices they need and want,” he said.
    SIDBI MSME Conclave 2024 |Register Now.
    ...more
    The Economic Times

    Stories you might be interested in