How Nugget Helped Zomato Reduce Support Cost by $10 Million+

Escalating support costs

Customer delight operations relied on 4,000+ outsourced agents, costing approximately US $20M*.


Tool sprawl and data silos

Operations were spread across 8 to 10 separate platforms (Zendesk, Freshdesk, Zoho, Salesforce, etc.), leading to fragmented data and limited insight.


Underperforming automation

Although approximately 60% of tickets were auto resolved (pre Gen AI era), bot CSAT lagged significantly behind that of human agents.

* Proportionated to current order numbers at historical cost of resolve

AI Agents for Chat

A multilingual bot that detects user intent in real-time, analyzes images (e.g., spills, tampered packaging), and triggers instant actions such as escalation, cancellation, or refund.


Agent Co-Pilot

Provides instant access to customer profiles, issue diagnosis, recommended resolutions, and auto generated chat summaries for every human handled ticket.


Unified omnichannel workspace

A single platform that routes and resolves every ticket, across chat, voice, social media, in app, and internal channels, replacing the fragmented legacy CX tools.

Image analysis

Nugget's vision system scans user submitted photos to identify wrong items, spills, damaged packaging, and more; enabling Zomato to automate up to 70% of refunds. It can also detect stock images, brand mismatches, and defect types.


Multi-modal LLM orchestration

For every user interaction, Nugget dynamically orchestrates the most cost efficient and high precision models, text, vision, or speech, by leveraging multiple large language models (LLMs) in parallel. This ensures maximum accuracy at the lowest AI cost per ticket in customer experience (CX).


100% Quality Control

Each and every AI chat is audited and analytics are provided real-time. This enabled Zomato to create automation logics by identifying patterns and ticket nature buckets.