Power your chatbot with the same Prefixbox search engine that’s on your online store to return relevant search results.

Leverage text-based, vector-based, and Hybrid search for data-driven search results to enable product search in Infobip Answers.

Prefixbox’s automated Hybrid Search Engine uses AI and machine learning to analyze queries and understand user intent to ensure shoppers are presented with relevant results that make shopping easy.

Use cases

• Product search in chat.

How it works?

• It exposes the Prefixbox Search API.
• It maps the following search index fields: product name, product URL, image, price, and description.
• It returns the number of products you specify.

Features

• Vector Search. Return the most relevant results by leveraging technology that understands the semantic relationships between search queries and products.
• Hybrid Search. Decipher natural language queries by leveraging both text-based and vector-based search combined with GPT technologies.
• Query Understanding. Leverages natural language processing (NLP) to decipher nuances of everyday language and speech.
• Dynamic Re-Ranking. Dynamic Re-Ranking analyzes user behavior for each search query and re-ranks the top 200 search results based on different parameters, such as Click Through Rate for each query.
• Synonym mining and management. Automatically apply the same synonym rules you have set up for your online store.
• Matching rules and tuning weights. Specify where you want queries to match in the product feed (name, description, attributes, brand, etc.) and apply specific weights to each product field to tweak results.
• Popularity score calculation. Ensure the most relevant and popular products appear first with accurate ranking calculations that consider product popularity trends and click feedback.

Find full search feature list here.

Prerequisites

• Prefixbox search integration
• Share up-to-date product catalog on a URL
• Add Prefixbox event logging to your Ecommerce website

Pricing

FREE