Himani Shah is the co-founder and CFO of Intello Labs, one of India's leading agri tech firms which provides quality assessment of food commodities using computer vision and AI. In this interview, Himani discusses the application of AI in the Indian agricultural sector and the firm′s plans for expansion into the US and South East Asia.
How are AI, computer vision disrupting the fresh food supply chain market in India?
Food and agriculture had been left behind despite the advancement of AI in some of the other sectors. In the food supply chains, quality control and its dependent decision-making are still conducted with traditional manual methods - globally, and more so in India. Even leading food companies often don't have visibility into the quality of food at procurement and at various stages throughout their supply chain. Prone to manual error, quality becomes subjective and inconsistent, especially in large organisations, leading to food loss and economic losses. Closer to food origination, farmers bear the direct burnt of lack of objective quality data through the loss of income. This is what computer vision is disrupting within the Indian food supply chains. The use of AI is helping transform transactions, quality management, logistics and distribution functions, enabling stakeholders to do more with less, in an objective and scalable manner and reducing food waste along the way.
What sets Intello apart in the agri tech sector Please provide some details on the company's flagship product Intello Track.
Intello Labs combines the tools of AI and data sciences to quality-related use cases with fresh produce. We are doing it in a way that is low-cost, versatile, mobile and highly scalable, being able to be plugged into various points in supply chains. In terms of usability, this is a great leap forward from both the manual or lab-based methods of determining quality.
Intello Track is an online quality monitoring platform that provides real-time visibility and calls-to-action across stages of a supply chain. While for quicker deployments it works with smartphones, more robust deployments involve image capturing through HD cameras mounted on ceiling or on shelf bots. It is appropriate for food aggregators, supermarkets and e-grocers. It enables upstream interventions too, providing decision support for the procurement department and quality checking at arrival and dispatch of warehouses and fulfilment centres.
According to government data, post-harvest losses are highest in the fruit and vegetable sector with as much as 16 per cent of produce going waste. What role can the advanced technologies play to mitigate food waste at the farm level?
For fruits and vegetables, it becomes difficult to monitor detailed crop health and growth status. In the absence of real-time data, farmers tend to apply inputs uniformly throughout the fields, which causes costs to rise and unnecessary amounts of fertilisers and pesticides to enter the food supply chain. Similarly, with no insights on crop maturing, a farmer typically harvests all of his fields at one go at the first sight of flowering anywhere in the field - not letting the entire crop properly mature. This leads to upstream food losses. Additionally, the lack of appropriate sorting capabilities also leads to food losses in the longer retail supply chains.
How many national and international retailers are you working with currently?
We are currently working with 4 of India's major retailers including Reliance Fresh, while piloting with several other multinational retail corporations & e-grocers globally. We are also working with several clients outside the retail sector.
What are the top two-three areas where the funding raised so far has been deployed?
The three broad areas with maximum fund utilisation have been technology (ie our proprietary model development), product development and sales. For each commodity, we develop a neural network algorithms that identify individual units of produce and detect contours on their surface. This task requires huge data collection and training across various machine learning models. Our repository of images, across commodities, runs into 40+ million. Speaking about product development, Intello has multiple product modules to address varied use cases across fresh food supply chains.
Intello is reportedly working on its global expansion plan with a focus on USA and South East Asia region. What are the criteria to choose markets/countries for expansion?
Our products have a global appeal as food supply chains the world over are non-digitised. To set immediate priorities, we studied the markets' respective pain points - food waste, customer quality, labour costs to quality, etc. Moreover, we analysed the level of competition, market access strategy and, of course, the revenue potential.
Based on the expansion plan, what would be the quantum of the funding requirement going ahead?
With the latest funding round, we are comfortably placed for the next couple of years. Given the Covid situation, our objective is to preserve cash and see through the tough times.