Many of us use quick commerce platforms almost every day. From groceries and snacks to even electronics, getting items delivered in minutes has become the new normal. But what many people are not aware of is the complex system working behind the scenes to make this possible. Recently, I got a chance to visit Zepto’s Mother Hub in Bengaluru, a massive facility where products from brands and suppliers first arrive before being sent to the company’s dark stores. Watching the operations behind the scenes made one thing clear: quick delivery is not just about speed, it is also about a well-coordinated operation powered by technology.
In an interaction with Digit, Zepto Co-Founder and President – Technology and Product, Kaivalya Vohra, explained how the company is using AI across its operations, why automation is about assisting people rather than replacing them, and how technology is helping everyone from customers to farmers.
When people think of AI in apps, they usually imagine product recommendations or a chatbot answering customer questions. Zepto uses AI for those features, but that’s only one part of the story. ‘AI is already a meaningful tool in quick commerce because it supports better decisions across the customer and operating experience,’ Vohra said.
According to him, AI works behind the scenes at almost every step of the shopping experience. ‘At Zepto, AI is used in practical ways across the customer journey: making search more relevant, improving product discovery, powering recommendations and supporting customer queries,’ he said.
One of Zepto’s biggest AI-powered tools is Zepto Atom, an in-house platform built for brands. Atom gives brands useful insights into what customers are searching for, which categories are growing and what people are buying. ‘It helps brands understand customer searches, category behaviour and purchase patterns on Zepto, so they can make more informed choices around product visibility, placement and campaign design,’ Vohra explained. ‘For a newer brand, that can mean reaching customers who are already looking for a relevant need, rather than relying only on broad-based advertising.’
The most interesting use of AI at Zepto isn’t inside the app at all. It starts much earlier, on farms. India’s agriculture has traditionally been supply-driven. Farmers usually decide what to grow based on weather conditions, previous crop cycles and local market trends. However, predicting what customers in cities will actually buy is much harder. This is where AI steps in.
Zepto’s demand forecasting models analyse buying patterns across cities, neighbourhoods, seasons, festivals and different price points. These insights help the company’s sourcing teams understand future demand more accurately.
‘For the farmer, the relevance is straightforward: better visibility into what is being bought, where demand is emerging, and the quantities required,’ Vohra said. ‘That supports more informed sourcing conversations, helps reduce avoidable mismatch between supply and demand, and can contribute to lower wastage in the fresh-produce chain.’
The company does not use traditional warehouse robots across its fulfilment network. Instead, it has developed its own automation systems, including Falcon and Neo, to solve specific operational challenges. These systems support workflows inside fulfilment centres by handling routine movement and repetitive processes. ‘The purpose is not technology for its own sake. It is to improve consistency and reduce dependence on repetitive, physically intensive activity,’ Vohra explained.
Whenever automation enters a workplace, concerns about jobs naturally arise. Vohra believes automation changes the nature of work rather than replacing workers. ‘Technology changes the nature of work, particularly in operations,’ he said.
Instead of spending hours on repetitive and physically demanding tasks, employees can take on roles involving quality control, exception management, systems support and maintenance. ‘The objective is to make work safer, more productive and more skill-led – not to treat automation as a substitute for people,’ Vohra explained.
Despite the growing role of automation, Vohra believes people are important. According to him, repetitive, high-volume and physically demanding tasks are ideal for automation because machines can perform them consistently. However, many jobs still require human judgment.
Fresh produce is a good example. Assessing the quality and freshness of fruits and vegetables cannot always be reduced to a fixed set of rules. Similarly, handling unusual situations, solving operational problems and making customer-related decisions still depend on people. ‘The best operating model is not humans versus machines,” Vohra said. “It is technology handling repeatable tasks at scale, while people focus on work that requires judgment, quality control and accountability.’
Quick commerce is no longer limited to groceries. Customers now buy smartphones, smartwatches and other expensive gadgets through these platforms. When asked about the company’s verification process for high-value electronics, Vohra said Zepto works with sellers to ensure products are sourced through authorised partners. The company also uses OTP-based delivery for expensive products, which allows customers to check the package before sharing the OTP.
‘Where a concern is raised, we review it with the relevant partners and stakeholders and take appropriate action based on the findings,’ he said. According to him, customer trust remains central to how Zepto handles these categories.
For most people, quick commerce begins when they open the app and ends when the delivery arrives. However, behind the scenes, a huge amount of technology is constantly working to make that experience possible. A single order requires the app, fulfilment centres and delivery network to stay connected in real time. ‘Most people see a delivery arrive in minutes. What they do not see is the operating system behind it,’ Vohra said. He added, ‘the combination of local operating complexity and technology is what makes quick commerce in India such an interesting innovation story.’
After spending time at Zepto’s Mother Hub, one thing became clear to me. Quick commerce is much more than delivering an order in a few minutes. Behind every delivery is a network of people, technology and AI working together. While customers only see the app and the delivery partner at their doorstep, there is a lot happening in the background to make that experience smooth.