Founded in 2017, Yulu is a technology-driven platform that provides micro-mobility.Operational in 5 cities across India, it is on a mission to make urban mobility in India seamless, shareable and sustainable1. Yulu started off with a smart lock enabled bicycle, Yulu Move, to solve the first mile, last mile and the short commute challenges in urban areas. Later in 2019, it launched a fleet of electric bikes, Yulu Miracle. In the newer initiatives, Yulu has launched Dex, an electric bike suited for gig economy workers.

We interviewed Jaideep Dhok, Head of Data Engineering & Amit Kumar, Lead Data Engineer for the Case Study. Jaideep heads all the data initiatives for Yulu which includes, Data Science, Data Analytics & Data Engineering. Amit leads the 12 member data team. The team has 4 data engineers, 2 data scientists & 6 data analysts. Together they work to build models & generate various reports & dashboards, which are used by various teams in Yulu. The operations team spread across 5 cities, Delhi, Mumbai, Bangalore, Bhubaneswar & Pune uses it to monitor on-field tasks. The product team uses reports & dashboards to track product releases, access the response, impact & other product metrics. The customer service & the strategy team is also aided by the data team for their data requirement. Most of these data assets are built and consumed using Sprinkle’s Data Integration & Analytics Platform.

In the initial days, the data team faced challenges on multiple fronts. The team was diligently building various solutions themselves and was also on the lookout for solutions that could provide them with quick results. As the operational data grew, the relational database used by the team became bulkier, and the team wanted to push data to a data warehouse. The team needed an end-to-end solution that could be integrated easily. A solution that could provide them with seamless data ingestion from various sources like relational databases, Kinesis, Kafka& CSVs. Help them to transform data, build models, flows, reports & dashboards for their teams briskly. The team wanted to build a complete data stack. The task at hand was also to build data models, like Demand Forecasting and Tracking Bikes in real-time. These models are business-critical and require the urgent attention of the team.
‍
Being at the centre of all data, the data team received many requests from other teams. Some of these teams who would consume the data assets created by the data teams lacked expertise in the query language. So, the team was also on the lookout for a self-serve, no-code kind of analytics platform to free up much of their time resources.

The team started out with configuring their data warehouse to their needs. The Sprinkle team got in touch and did a proof of concept with the data team, demonstrating various features and capabilities of the Sprinkle platform. On-site training for smooth onboarding was provided to the team members. Advanced data analytics models were built using the feature-rich Sprinkle platform. The streaming events data from the bikes were used to build real-time data models to help the operations team perform maintenance and track bikes. The data forecasting model built help the operations team to plan the availability of bikes in different zones. The team also built multiple customised dashboards for specific needs.

Major analysis that empowers the key business tasks is done using Sprinkle. Over 20+ analysts of the central operations team located across 5 cities, use sprinkle reports &dashboards to help operations. With more than 50 connectors that are available on Sprinkle for the ingestion, the team ingests data with ease. The team has also observed an increase in the productivity of the team members. The no-code way of building models is being extensively used.They have considerably reduced the time taken to develop models by migrating to theSprinkle No-Code way from the traditional way of writing the code.