AI recommendations and personalisation for online retail
A recommendation and personalisation engine across a 1,80,000-SKU catalogue, with AI cart recovery and 24/7 chat support.

The challenge
An online retailer with 2.4 million monthly visitors showed every shopper the same generic product display. Conversion was stuck at 1.2%, against an industry average of 3 to 4%, and the merchandising team could not keep up with a catalogue of 1,80,000 SKUs. Cart abandonment had reached 78%, costing an estimated Rs. 4.2 Cr in revenue every month.
Our solution
We built a collaborative filtering recommendation engine across the full catalogue and a behavioural segmentation model from clickstream and purchase history, so every shopper sees products that match their intent. AI-triggered cart recovery emails use dynamic discount logic, and a conversational chatbot handles support and returns around the clock.
What we delivered
- Collaborative filtering recommendations on 180K SKUs
- Behavioural segmentation from clickstream and purchases
- AI cart abandonment emails with dynamic discounts
- Conversational AI chatbot for support and returns
“The recommendation engine alone paid for the entire project in 6 weeks. Our customers now feel understood.”
Key results
Results measured with the client. The client’s name is withheld at their request.
Services used
Technology stack
More case studies
All case studies
An AI chat assistant for personalised revision and mock exams
A chat-based AI assistant that helps students study daily, revise, sit mock exams and get support in many languages, on any device.
Read case study
Adaptive learning and assessment for financial education
An AI learning engine for accounting students, CA aspirants and finance professionals, with adaptive study paths, smart practice and mock exams.
Read case study
An AI-powered ERP for a garment manufacturer
One AI-integrated ERP for production, inventory, procurement and dispatch at a factory making over 2 lakh garments a month.
Read case studyYour project could be next.
Share what you want to build and we will reply within one business day with questions and a suggested next step.