DataArt & Rappi: Three Pillars of Strong Technological Partnership

Engineering partner:DataArt
Location:LATAM
Industry:Travel Tech Solutions
Services:Legacy Modernization

Client

Rappi is a multi-vertical super-app that offers food, groceries, e-commerce, pharmacy, travel, and payment services in 9 countries in Latin America. One of Rappi’s key verticals is Rappi Travel, an online travel platform for booking trips, aggregating flight offers, and more. This vertical aims to continue to add value to the lives of Rappi users by providing multiple integrated services within the super-app ecosystem.

Challenge

Against a backdrop of increased demand for travel and constantly fluctuating travel conditions, Rappi recognized the need to improve the app’s search features and capabilities for its customers. Its Airport & City Search Engine presented an opportunity to enhance accuracy and performance while building a curated, seamless search experience. Goals: 1. Improve user experience by returning faster and more accurate results 2. Leverage sales by delivering the proper search response to the visitor 3. Decrease maintenance costs by introducing a new microservices architecture

Solution

DataArt’s vision: Improving in-app search without the operational overhead and creating a tolerant and relevant process that offers users an out-of-the-box search tool inside the app. DataArt introduced a new search engine architecture based on microservices and custom indexing, using Amazon OpenSearch Service at its core. The solution resembled a real-life search experience with searching functionalities of unrivaled relevance.

Highlights

  • Improved indexing and search performance by 10X
  • Increased maintainability, scalability, and reliability

Benefits

  • Improved conversion rate for search by 5% for locations not covered by popular destinations
  • Reduced cost-to-serve by optimizing marketing performance and costs
  • Increased engagement and customer satisfaction with a rich search experience
  • Decreased maintenance costs
Indexer features
  • Ability to customize Amazon OpenSearch Service indexes, including mappings and analyzers (special characters, synonyms, and stop-words)
  • Seed and update curations in Amazon DocumentDB
  • Load/update Amazon OpenSearch Service indexes with curated cities and airports using destination's API as a source
Search features
  • Search functionality based on a search query, language, and country
  • Enable field weights and scoring functions to optimize search results based on language, country, and custom score provided by curations

Technology Stack

  • AWS
  • Amazon OpenSearch Service
  • Amazon DocumentDB