The Challenge
E-commerce businesses face enormous pressure managing product-related customer queries. For Omlet, a leading online pet products retailer, the challenge was particularly acute—thousands of daily enquiries spanning product specifications, assembly guidance, compatibility questions, and troubleshooting requests. Each query required accurate, product-specific knowledge that traditional support processes struggled to deliver consistently.
The scale of the challenge meant customer service teams spent hours researching answers across scattered documentation, previous correspondence, and product manuals. Response times stretched from minutes to hours, customer satisfaction suffered, and operational costs continued to climb. With a diverse product catalogue including complex items like automatic chicken coop doors and modular pet housing systems, every query demanded specialist knowledge that was difficult to maintain across a growing support team.
Our Approach
New Gradient developed a sophisticated Retrieval-Augmented Generation (RAG) system that transforms how Omlet's customer service operates. The solution ingests and indexes the company's entire knowledge base—product documentation, historical customer correspondence, assembly instructions, FAQs, and technical specifications—creating a comprehensive, searchable repository of product expertise.
At the core of our system, vector embeddings convert this vast documentation into a format that enables semantic search, allowing the system to understand the meaning behind customer queries rather than relying on keyword matching alone. When a customer asks about compatibility between products or troubleshooting a specific issue, our retrieval system identifies the most relevant documentation across thousands of sources in milliseconds.
We then fine-tuned large language models specifically for Omlet's customer service context, training them to generate responses that match the company's tone and style while maintaining factual accuracy. The models learned from years of successful customer interactions, understanding not just what information to provide but how to communicate it effectively. Crucially, the RAG architecture ensures responses are grounded in actual product documentation rather than generated from the model's general knowledge, virtually eliminating hallucination.
The Outcome
The transformation in customer service efficiency has been remarkable. Response times dropped from hours to seconds, with the system providing accurate, contextually appropriate answers to the vast majority of incoming queries. Customer service agents now receive AI-generated draft responses that they can review, refine if needed, and send—multiplying their effective capacity while maintaining the human touch for complex or sensitive enquiries.
The 94% accuracy rate means customer service staff can trust the system's suggestions, dramatically reducing the time spent researching answers. For straightforward queries—product dimensions, compatibility checks, standard troubleshooting—the AI handles the heavy lifting entirely. This frees human agents to focus on complex cases that genuinely require their expertise and judgement, improving job satisfaction alongside operational metrics.
Beyond immediate efficiency gains, the system continues to learn. As new products launch and documentation updates, the knowledge base expands automatically. Customer interactions help identify gaps in documentation and common points of confusion, creating a feedback loop that improves both the AI system and the underlying product information. The result is customer service that gets smarter over time, scaling effortlessly with business growth.