Achieved 90% accuracy in order forecasting: Rebel Foods

Aug 21, 2020

It has been Data science model entwined with Machine Learning algorithms that made it easier for Rebel Foods to forecast demand and order with a 90% accuracy level to state analytically.

The Sequoia-backed company, Rebel foods, is a universal parent to cloud kitchen brands like Behrouz Biryani, Faasos, Mandarin Oak, SLAY, Sweet Truth, The Good Bowl and more. The company serves extensively with million orders in a month with over 3000 restaurants globally and 327 kitchens in India.

Being a new-age company with technology rooted in its DNA, Rebel food aims to deliver a seamless customer experience with no inventory wastage in the kitchens.

“As a company, it is in our DNA to make data-driven decisions. Be it solving a business or a customer problem or launching a new product, data has always been the deciding force for us,” said Amit K. Gupta, CTO, Rebel Foods.

“There’s a thin line between running businesses as usual and making a compromise, between a targeted attack on your company and the moment you first discover the breach, and between being proactive and reactive to such breaches. The only way to be on the right side of the line is by using ‘time’ to your advantage. Predictive advantage is a measurement of the time difference between the creation of an AI cyber-security model, and the first time a threat is blocked by that same model in the wild…” added Gupta.

As of today, Rebel foods has collected and organized data from past 5 years of orders from 35 various cities of India. As the company is expanding its business globally, simultaneously it is also collecting data from those cities as well.

“We consolidate all the order data, behaviour data, and the questions we have been asked by customers and interaction data. On top of this, we run a data science ML algorithm to do to the demand forecasting and on the basis of the same, we do the inventory planning in each and every kitchen,” explained Gupta.

Since the kitchens have a break-up of inventory for every dish they prepare and serve, the company can plan beforehand how much will be required on a particular day in the kitchen. Owing to such practice of pre-planning, Rebel foods has minimized its food wastage drastically.

“We have pretty good accuracy on these forecast owing to the amount of data we have. We have achieved 90 % accurate in demand and order forecasting at every kitchen level,” Gupta said.

“Wastage for us is very minuscule. Because we rely a lot on data to predict these and it’s working beautifully in that part,” added Gupta.

The technology teams have worked to have multiple systems integrated at Rebel Foods, Gupta believes in solving the problem statements faster in the food industry, an industry which has not been disrupted in the last 500 years.

“We have to build a lot of systems in the house to give us speed in customizing those systems as per our need. The inventory taking system in the kitchens, for example, is a mix of automated and manual stuff. Automated because we use technologies like a barcode scanner. The machine is integrated with our software component where it automatically takes in the data into the system when you scan the product and there is some manual input also required where dashboards are provided to kitchen staff with a lot of checks and approval where they can put the inventory data in the system,” Gupta explained.

He went on to add, “Everything has to be put in the consolidated system and there are approval mechanisms and checks and balances. At the end of the day too, the kitchen staff scans the barcode and the systems get to know how much inventory is still left at the kitchen level”.

“Our audience, the new generation is accepting technological advancements faster than ever. They expect us to provide them with a seamless experience through these advanced technologies. And the companies which do not embrace this change will ultimately be left far behind” he concluded.

Leave a Reply

Your email address will not be published. Required fields are marked *