Why Zomato is cutting up to 300 support jobs as AI automation reaches 85%

Zomato is shutting its Hyderabad customer-support centre and eliminating an estimated 250 to 300 customer-service roles as parent company Eternal Limited (NSE: ETERNAL; BSE: 543320) increasingly shifts routine support work toward artificial intelligence, automation and external service providers. The Economic Times first reported on August 31 that approximately 250 positions were affected as Zomato consolidated its in-house support operations at its Gurugram headquarters, while subsequent reporting put the potential reduction at 250 to 300 roles.

An important employment distinction is that the affected Hyderabad workers are employed through third-party service providers rather than sitting directly on Zomato’s payroll, according to Business Standard. The restructuring therefore represents a contraction in the labour supporting Zomato’s customer-service operation without necessarily producing the same direct employee headcount reduction that would appear if Eternal itself terminated 250 permanent employees. The operational consequence is nevertheless significant because an entire Hyderabad support centre is being closed while customer-service activity becomes increasingly concentrated in Gurugram and technology-driven channels.

The numbers behind that technology shift explain why this is more than a routine location consolidation. Business Standard reported, citing company data, that Eternal’s Nugget AI customer-support platform has handled more than 2.2 billion conversations, provides automation coverage across approximately 85% of interactions and maintains 99.9% service availability. The same reporting said Nugget has helped reduce annual customer-support costs by around 55%, from approximately $20 million to $9 million, while handling more than 15 million conversations each month.

Why is Zomato shutting its Hyderabad customer-support centre now?

The Hyderabad closure follows roughly six months of changes to Zomato’s customer-support model, during which the company has increased both automation and the proportion of work handled through external partners. The Economic Times reported that Zomato is consolidating the remaining in-house operation at its Gurugram headquarters, effectively reducing the need to maintain a second large customer-support location in Hyderabad.

That makes the restructuring fundamentally different from a conventional revenue-driven layoff. Zomato’s parent company is not shrinking because customer activity has collapsed. Eternal reported strong overall growth in its latest quarter, while the Zomato platform itself continues to serve more than 27 million monthly transacting customers and more than 300,000 restaurant partners. The issue is that supporting those customers increasingly requires fewer human interactions per transaction as automated systems handle more routine queries.

The economics of running two support centres therefore change once the majority of customer conversations no longer require a traditional call-centre workflow. A physical centre carrying management, facilities and service-provider costs becomes harder to justify if AI agents can resolve most repetitive customer problems and more complicated cases can be routed to a smaller centralised team.

The Hyderabad closure consequently looks less like an isolated office decision and more like evidence of how AI can alter the physical footprint of a consumer-internet company even when the underlying business continues expanding.

How significant is Nugget AI’s reported 55% reduction in customer-support costs?

A reduction from approximately $20 million to $9 million would imply about $11 million of annual support-cost savings, which is significant even for a business the size of Eternal because customer service is traditionally a recurring operating expense that scales with transaction volumes. If order volumes rise while support costs remain relatively stable or decline, each incremental transaction becomes easier to service economically.

The more striking figure is the reported 85% automation coverage. That suggests automation is no longer being used merely to answer a narrow category of frequently asked questions. It has moved toward becoming the default layer handling customer interactions, with humans increasingly concentrated on exceptions, escalations and cases where automated resolution is insufficient.

Business Standard also reported that Nugget handles more than 15 million conversations per month. At that scale, even small reductions in the cost per interaction can produce meaningful annual savings because the technology is applied repeatedly across millions of customer contacts rather than being deployed as a one-time productivity tool.

That is what makes the Hyderabad restructuring commercially important. Many companies describe artificial intelligence as a future efficiency opportunity. Zomato appears to be demonstrating a measurable cost outcome alongside an actual reduction in the amount of human support infrastructure it requires.

Is artificial intelligence directly responsible for the 250–300 affected support roles?

AI is clearly one of the factors changing Zomato’s support model, but it would be too simplistic to attribute every affected role exclusively to artificial intelligence. Reporting indicates that Zomato has also expanded outsourcing to external service providers and is geographically consolidating its own support operation in Gurugram, meaning several organisational changes are occurring simultaneously.

The stronger conclusion is that automation has changed the economics of customer support enough to enable a smaller and more centralised operating model. When AI handles a larger proportion of conversations, the company can reduce the volume of work allocated to human agents, while outsourcing provides flexibility for whatever human workload remains.

This distinction matters because AI-driven workforce change does not always appear as a company replacing a particular employee with a particular software agent. More often, automation gradually reduces the number of cases requiring human attention, after which management consolidates teams, does not refill positions or closes locations whose previous staffing level is no longer economically necessary.

Zomato’s Hyderabad decision appears to follow that broader pattern. The technology changes the workload first; the organisational footprint adjusts afterward.

Why does the restructuring stand out when Eternal’s revenue is growing so quickly?

Eternal’s latest financial performance shows that the support reduction is occurring against a backdrop of expansion rather than corporate contraction. In the June quarter, consolidated adjusted revenue increased 173% year over year to ₹20,648 crore, while consolidated adjusted EBITDA increased 223% to ₹555 crore. Business-to-consumer net order value grew 54% to ₹31,120 crore.

On a reported accounting basis, Eternal generated consolidated revenue of ₹20,211 crore and net profit of ₹92 crore, compared with ₹7,167 crore of revenue and ₹25 crore of profit a year earlier. The company’s profit therefore increased more than threefold year over year, although it declined sequentially from ₹174 crore in the preceding quarter.

These numbers change the interpretation of the layoffs. Eternal is not reducing customer-service capacity simply because revenue has fallen and management needs emergency savings. Instead, technology appears to be allowing certain operating costs to grow far more slowly than the customer and transaction base.

That is potentially a much more powerful business model effect. If customer volumes increase while the cost of resolving each support interaction falls, operating leverage can improve even before the company makes major changes to pricing or commissions.

How large is Zomato’s customer operation that Nugget now supports?

Zomato reported more than 27 million average monthly transacting customers during the first quarter of fiscal 2027, supported by more than 300,000 average monthly active restaurant partners. Eternal also says more than 600,000 delivery partners earn through the Zomato ecosystem.

That scale creates an enormous support challenge because food delivery produces countless situations requiring customer communication: incorrect items, delayed orders, cancellations, refunds, payment problems, restaurant availability, delivery-partner issues and complaints over product quality. Historically, handling those interactions at scale could require large teams of support agents.

AI changes the relationship between transaction growth and employment growth if the system can reliably classify problems, retrieve order information and execute standard resolutions without intervention from a human agent. The company can then process more orders without increasing support headcount at the same rate.

The economic attraction is obvious, but execution quality remains important. A customer whose food order has gone wrong often wants an immediate resolution rather than a generic automated response. If automation reduces costs but increases frustration or produces incorrect outcomes, short-term support savings could eventually damage customer retention.

Zomato therefore needs to optimise two variables simultaneously: cost per interaction and quality of resolution.

Why could Nugget become more valuable outside Zomato than inside it?

Eternal is not treating Nugget purely as internal infrastructure. Its privacy documentation describes Nugget as an AI-agent service offered to corporate clients, allowing businesses to connect their own data sources and use AI agents for customer-support and related workflows. The platform can integrate with third-party large-language-model providers while operating on information supplied by individual clients.

That potentially changes the economics of the technology investment. Software developed initially to reduce Zomato’s own support costs can become an external product generating revenue from other companies facing similar customer-service challenges.

The Hyderabad restructuring therefore provides an unusually tangible proof point for that commercial pitch. If Eternal can credibly demonstrate that its own AI platform helped reduce customer-support costs by roughly 55% while operating across millions of monthly conversations, prospective enterprise customers have a concrete internal use case rather than merely a product demonstration.

That does not automatically mean external customers will achieve comparable savings. Industries have different support complexity, regulatory requirements and customer expectations, and an AI system effective for food-delivery refunds may require substantial adaptation for banking, healthcare or other sensitive sectors.

Still, successful internal deployment gives Nugget a stronger commercial story than an AI product built without equivalent production-scale experience.

Could more customer-support jobs disappear if automation keeps improving?

The Hyderabad closure demonstrates that further efficiency gains could have workforce consequences, but there is currently no disclosed company-wide target for additional customer-support reductions. It would therefore be speculative to extrapolate the current 250–300 affected roles into a much larger future layoff figure.

The underlying direction is easier to identify. If automation coverage moves beyond the reported 85% while the quality of automated resolution remains acceptable, the number of conversations requiring human intervention could decline further even as total transaction volumes increase.

Human support is unlikely to disappear entirely because unusual complaints, safety issues, payment disputes and emotionally sensitive situations often require judgment that standard automated workflows cannot consistently provide. The role of human agents may instead shift toward more complex exception handling while AI handles routine cases.

That transition can still reduce total employment because complex cases represent a smaller percentage of overall volume. The workforce implication depends less on whether AI becomes capable of resolving every possible query and more on whether it resolves enough routine queries that fewer agents are needed for the remainder.

What does Eternal’s share-price performance say about investor sentiment?

Eternal shares closed at ₹327.70 on August 31, almost unchanged from ₹327.75 in the preceding session. Over the five trading sessions beginning August 24, the stock was also essentially flat, moving from ₹328.00 to ₹327.70. That suggests the Hyderabad restructuring did not produce an obvious negative market reaction when the story emerged.

The one-month trend is substantially stronger. Eternal closed at ₹302.35 on July 31, meaning the August 31 price represented an increase of about 8.4%. Its 52-week range stands at approximately ₹212.55 to ₹368.40, leaving the latest closing price around 11% below the annual high and approximately 54% above the low.

Investor interest is being driven by much larger factors than a few hundred support roles, particularly the rapid expansion and improving economics of Blinkit. Reuters reported that Blinkit’s net order value increased 86% year over year during the June quarter and that adjusted EBITDA margin improved to 0.6%, with management expecting long-term margins toward the upper end of its previously indicated 5% to 6% range.

The support restructuring nevertheless fits the same wider investor narrative. Eternal is attempting to prove that rapid growth can increasingly coexist with operating efficiency, whether through better quick-commerce economics or lower costs in mature functions such as customer service.

Why does the Hyderabad closure matter beyond 250–300 individual roles?

The bigger significance lies in what it demonstrates about the labour intensity of digital consumer businesses. Online delivery platforms once needed customer-service workforces capable of expanding alongside transaction volumes because every incremental million orders created additional complaints, refunds and queries requiring human handling.

AI changes that assumption.

If Eternal can support a substantially larger transaction base while spending approximately $9 million annually on customer support rather than about $20 million, the company gains operating leverage that would have been difficult to achieve simply through wage negotiations or conventional call-centre optimisation.

There is also a geographic consequence. Hyderabad has become an important Indian centre for technology and business-process employment partly because companies can centralise large customer-support and back-office teams there. If AI increasingly reduces the number of people required for standard service interactions, future restructuring could affect not only individual employers but the composition of business-process work across major Indian cities.

That does not mean support employment disappears. Companies may continue employing specialists, escalation teams, quality analysts and technical experts while creating new roles involved in training, monitoring and improving AI systems. The mix of work changes, however, and routine interaction-processing jobs appear especially exposed.

What should investors and employees watch after Zomato’s Hyderabad support closure?

The first number to watch is Nugget’s automation coverage. The reported 85% level is already high, so further increases would indicate that AI is moving deeper into more complicated interactions rather than simply automating easy queries. The second is customer-support cost, where the reported decline from approximately $20 million to $9 million provides a measurable benchmark for whether automation continues producing financial benefits.

Customer experience is equally important. Cost savings become much less attractive if complaints increase, refunds are mishandled or customers find it difficult to reach human support when genuinely necessary. Eternal’s ability to preserve service quality while reducing human involvement will determine whether the automation model can scale sustainably.

The third issue is Nugget’s external commercial adoption. If companies outside Eternal begin paying to use the technology, an internal cost-reduction tool could evolve into another software revenue stream. Successful externalisation would make the Hyderabad restructuring part of a broader technology story rather than simply a call-centre closure.

Finally, investors should watch whether similar organisational changes emerge elsewhere inside Eternal. A company demonstrating significant productivity gains in one function may naturally begin examining whether the same AI-agent architecture can automate processes in merchant support, internal operations, logistics administration or other labour-intensive workflows.

Zomato’s decision therefore carries more significance than its 250–300 affected roles might initially suggest. Eternal’s consolidated adjusted revenue increased 173% in the latest quarter, yet the customer-support infrastructure needed to sustain that growth is becoming smaller and more automated. That is precisely the operating leverage technology companies increasingly want AI to deliver.

For the Hyderabad workers affected through third-party service providers, the consequence is immediate employment disruption. For Eternal, the strategic question is larger: whether Nugget can continue lowering support costs without degrading customer experience and whether the same technology can become commercially valuable outside Zomato.

If it can, the most consequential number in this story may ultimately not be the 250–300 roles disappearing in Hyderabad. It may be the 85% of support interactions that apparently no longer need the traditional human-service model that created many of those roles in the first place.