Chope: Using Google Cloud to help people discover new restaurants, cuisines, and dishes

About Chope

Chope is a dining platform that operates in Singapore, Hong Kong, Shanghai, Bangkok, Phuket, Jakarta, and Bali. Approximately 3,000 restaurant partners and 1.5 million people use Chope’s websites and applications to discover restaurants, make bookings, and find dining deals.

Industries: Travel & Hospitality
Location: Singapore

Tell us your challenge. We're here to help.

Contact us

With Google Cloud, Chope quickly processes, analyzes, and reports on 775,000 records per day, helps users discover restaurants around them, and enables users to search for restaurants that serve dishes they crave.

Google Cloud Results

  • Enables the business to derive timely analysis and insights from 775,000 records per day
  • Allows the business to add restaurant discovery functionality to its websites and applications
  • Gave the business the ability to launch a new search function in just one month

Supports about 3,000 restaurants and 1.5 million users

East Asia is home to a wide range of cuisines and restaurants, making dining out an exciting experience for locals and tourists alike. In 2011, Singapore-based entrepreneur Arrif Ziaudeen and his business partners saw an opportunity to bring the convenience of online booking to the region. Seven years later Ziaudeen’s business, Chope, has expanded to Singapore, Hong Kong, Bangkok and Phuket in Thailand, Shanghai in China, and Jakarta and Bali in Indonesia. Today, approximately 3,000 restaurant partners and about 1.5 million people use Chope’s websites, applications, and various Facebook and restaurant website plug-ins to better their dining experience.

However, Chope’s services are not limited to restaurant bookings. Diners can earn loyalty Chope-Dollars that can be redeemed for cash vouchers at participating restaurants as well as other rewards. Chope also provides applications and services that enable restaurant owners to run their businesses more effectively, including a Table Management System that allows for efficient assignment of tables, the ability to seat and depart guests, and optimization of table turnaround times.

An innovation mindset

Ziaudeen’s innovation mindset helps ensure Chope is constantly improving the experience it provides to restaurants and diners.

After initially licensing software to run its operations, Chope moved its data, applications, and services to the cloud but still found it needed stronger data analysis and reporting capabilities.

“We were running a transactional database on our incumbent cloud provider and copied the information across to other software for analysis and reporting,” says Raymond Ong, Product Manager, Data & Logic at Chope. “However, the records were often duplicated or incorporated errors, and the process was slow and cumbersome. This presented a significant obstacle to obtaining timely, comprehensive insights to improve our decision making.”

Chope resolved to improve its analysis and reporting and began seeking a solution that would meet these needs without disrupting the business or its services.

“We received considerable support from Google in making the move and establishing a robust data warehouse and data pipeline. Google helped us identify solutions that could meet our business and technical requirements and provided guidance and feedback on our architecture.”

Raymond Ong, Product Manager, Data & Logic, Chope

BigQuery is easy to use

“We explored our options and the fast query speeds of BigQuery really prompted us to pursue it as our analytics data warehouse,” recalls Ong. “We found the product extremely easy to use while meeting all our speed and accuracy requirements.”

The Chope technology team was also impressed by the opportunities presented by real-time messaging service Cloud Pub/Sub as well as the data wrangling capabilities offered by Cloud Dataprep.

As a result, Chope decided to start using BigQuery and a range of other Google Cloud services. The business started the migration late 2016 and continues to operate a multi-cloud architecture.

“We received considerable support from Google in making the move and establishing a robust data warehouse and data pipeline,” says Ong. “Google helped us identify solutions that could meet our business and technical requirements and provided guidance and feedback on our architecture.”

Chope is using Cloud Dataflow to transform and enrich data from sources, such as its other cloud provider and a variety of marketing and event tracking platforms, and Looker Studio to prepare interactive reports and dashboards. Cloud Storage provides staging and data lake services.

Embracing machine learning

The business uses machine learning models to improve search and personalization capabilities for its services. Currently its machine learning and related data-processing workloads run on Kubernetes Engine clusters, and Chope plans to start utilizing Cloud Machine Learning Engine to run Tensorflow machine learning training applications as well.

Chope had initially tapped on Compute Engine to run its own Apache Airflow infrastructure, but later saw an opportunity to use Cloud Composer, then in alpha testing, to manage Apache Airflow workflow creation and scheduling instead. “Using this managed service means we do not have to dedicate time and resources to our own Apache Airflow infrastructure,” Ong explains. “We were excited to learn about Cloud Composer and to be invited into the alpha program.”

“Google Cloud Platform enabled us to embark on projects that a lack of technical resources would otherwise have stopped us from attempting.”

Raymond Ong, Product Manager, Data & Logic, Chope

Finally, while Chope’s use of Cloud Dataprep has tapered off after stabilizing its extract, transform, and load processes on Apache Airflow with Cloud Composer, the business still finds value in the product. “Cloud Dataprep is still a great tool for us to explore new unstructured data before undertaking any engineering work,” says Ong. “Its interface gives us an interactive way of visualizing data transformation and provides rapid feedback on transformation outputs. It also highlights missing or mismatched values in a dataset which makes it easy for us to draft out a transformation pipeline for a dataset.”

Completing projects with limited resources

“Google Cloud Platform enabled us to embark on projects that a lack of technical resources would otherwise have stopped us from attempting,” says Ong. “Furthermore, we can move faster. There is no way our two-person team would have been able to build our data warehouse as quickly as we did without the ease of use of Google Cloud Platform services.”

The Google Cloud Platform architecture has allowed Chope effectively utilize its data – approximately 775,000 records per day from structured and unstructured data sources – for rapid analysis and reporting. This has enabled the business to make more effective, timely decisions that improve its performance and enrich customers’ experiences with its websites and applications.

“We didn’t have much experience using Kubernetes to run containerized applications. With Kubernetes Engine, we were able to quickly learn the basics. This showcased how Google Cloud Platform made us more agile and able to adopt new technologies much easier.”

Raymond Ong, Product Manager, Data & Logic, Chope

Google Cloud APIs improve experience

Chope also uses Cloud APIs to expand on basic restaurant name searches and build features that enable users to discover exciting new restaurants. “We noticed that a lot of people liked to search for restaurants within certain areas or near landmarks,” comments Ong. “For example, they may be looking for a Japanese restaurant near a train station. By using Google Places information from the Google Maps API, people can select precise locations from a list of suggested results to find restaurants nearby.”

New search function rolled out in one month

Using Kubernetes Engine, Chope was able to roll out its new search function in just one month and seamlessly scale out the number of supporting virtual machines instances to help ensure query times met customer expectations. “We didn’t have much experience using Kubernetes to run containerized applications,” Ong says. “With Kubernetes Engine, we were able to quickly learn the basics. This showcased how Google Cloud Platform made us more agile and able to adopt new technologies much easier.”

Chope also uses Cloud Translation API to support Chinese-language searches for particular restaurant dishes. “Our research found the Cloud Translation API was the most accurate in translating Chinese search terms to English before being sent to our search engine,” Ong recalls.

With Google Cloud, Chope is well positioned to improve the depth and quality of its services – and expand into new markets.

Tell us your challenge. We're here to help.

Contact us

About Chope

Chope is a dining platform that operates in Singapore, Hong Kong, Shanghai, Bangkok, Phuket, Jakarta, and Bali. Approximately 3,000 restaurant partners and 1.5 million people use Chope’s websites and applications to discover restaurants, make bookings, and find dining deals.

Industries: Travel & Hospitality
Location: Singapore