With a fleet of more than 10,000 electric vehicles and a customer base of over 3 million users, Bengaluru-based Yulu is one of the largest shared electric mobility companies to revolutionise urban mobility in the country. Competing with Bounce, VOGO, and Ola Electric, the company provides first and last-mile connectivity to commuters. Also, it enables short-mile delivery of goods with their Urban mobility-as-a-service (UMaaS) platform. 

Yulu has a presence in Bengaluru, Delhi NCR, and Mumbai, and its technology-driven mobility platform utilises IoT, ML and AI for real-time demand-supply forecasting and efficient operations. In addition to this, the company has created the largest network of battery charging and swapping stations called Max Network to remove range anxiety and increase the adoption of electric mobility in India. 

All its electric vehicles come with embedded devices, sensors, motor controllers and battery management systems that help them track their e-bikes across the country and manage their operations seamlessly via consumer and in-house apps. 

The core of everything Yulu does today is driven by its data science team – from streamlining the operations of its e-bikes to tracking the conditions of vehicles across locations, alerting users about the battery level, forecasting demand and supply of vehicles, understanding consumer behaviour, and offering rental services to online food delivery partners, etc. 

In an interview with AIM, Jaideep Datta Dhok, head of data science at Yulu, shared some of the game-changing solutions that its data science team has delivered across departments, areas and last-mile delivery ecosystem, alongside their expansion plans to hire more data science talent in the coming months.  

Team structure 

Yulu has about ten people in the data science team and about 50+ people across departments. The data science team is flexible and subdivided into three verticals – 

• The data engineering team takes care of setting up the data platform, where the team is responsible for figuring out tools, technologies, cloud, visualisation, platforms, etc. 
• Next is the data analyst team: Here, the team is responsible for supporting the decisions for all the verticals of business – from founder to senior management roles. They are also responsible for understanding the data about the customers, operation process, tracking ever-evolving customer behaviour, etc.  
• Finally, the data science team is responsible for working on predictive modelling, where they forecast demand at different locations, battery supply, battery performance, etc. 

On a hiring spree…

“We are hiring in all three teams. By and large, we have some specialised requirements like data science, but we are focussing on hiring people who are interested in our ecosystem,” said Dhok. 

He said this domain is a bit different even if you look at similar companies operating in space – say, for example – they don’t own the fleet that their partners are using. So, this brings a lot of challenges, and we need to keep track of all aspects, including spare parts, maintenance, bike conditions during different seasons, operations, customer behaviour, etc. At the same time, they have to build data models where the data volume is huge, not only in terms of quantity but the complexity of developing such a solution in the ecomobility space. 

Dhok said, if you look at it, the data is rapidly changing not only on the customer side but also on the operating side, where the staff has to work on maintaining the fleet and ensuring all the e-bikes are in working condition. For instance, even on the maintenance front, Yulu tracks how much time the e-bike takes to get fixed/repaired, manage spare parts, charge the battery, etc., and optimise the process seamlessly without hindering the experience for end-customers. 

“We are looking for people who are not only interested but also curious and passionate about adding value to the entire ecosystem,” shared Dhok. 

Interview process 

Dhok said that for hiring, they usually have an exploratory call with the candidate, where they look at the experience of the candidate and then figure out if the claim in their CV is substantial and true to the best of their ability, alongside understanding the projects they have worked on previously. 

Once this is done, the candidate is escalated to the next stage of interview rounds. Here, they are either given certain problems to solve or explain why certain models and tools are used to derive the desired outcome. 

“We just want to know whatever you have put on the CV, you are able to explain,” said Dhok.  

He said, for instance, if you have written XGBoost on your CV, we would expect the candidate to know why they used it and how the performance was compared to other models. They are also keen on previous experiments, various challenges encountered in the process, how you solved them, etc. 

For freshers who do not have experience, the company does basic programming checks and does not dive into advanced algorithms. Yulu believes that the candidate gets to learn on the job and grow. However, when hiring for senior roles, the interview might get cumbersome. They need to know the complete lifecycle of deploying the models and develop new methodologies, alongside other qualities where they can mentor the team. 

Dos & don’ts

“Recently, I have seen a trend in candidates – say they learn some tools like Pandas or Spark – but lack the basic understanding of programming concepts,” said Dhok, highlighting some of the red flags. 

So, if you are planning to attend Yulu’s interview, make sure to brush on the basics skills instead of the tools they have been using lately. 

Work culture 

Yulu’s head of data science said that the company follows an open work culture across departments. “I would say, we give honest feedback… If you look, we don’t shy away from saying that there is a problem in which their tech model or data science model needs to be fixed,” shared Dhok. He said that the team is very open in expressing and taking feedback every day.  

The data team works very closely with both the business and tech teams. Within the data science team, there are no specific assigned duties, i.e., data engineers can take up analytics related problems from time to time to develop a better sense of the business. At the same time, data analysts can feel free to work on data engineering-related projects. “Freedom is always there,” said Dhok, stating that the tech team itself is quite flat and has a very friendly culture within the organisation.  

What sets Yulu’s data science team apart is it gives ownership to all its employees to work on challenging projects, along with the platform to have an open discussion/dialogue with the teams freely. In other words, Yulu has no hierarchy, flexible working hours, flexible leave policy and health and wellness benefits, among others. 

“I think most people really care about the problems that they are trying to solve. Also, the tech that we are trying to build is a bit different from other companies. And that brings in lots of interested people,” added Dhok. 

Why should you join Yulu? 

“It is one of the unique companies in India, working in this domain. We have a very data-driven culture – from founders to senior management levels – they know what data points should be looked at, and that is a very unique thing.”

Further, sharing his experience, he said: I have worked in many companies in my career at different stages of data adoption – there is so much struggle to convince the management to use the data – but, at Yulu, that is not the case, everybody looks at data, where they are constantly looking to optimise or add value in everything they do. “That is absolutely amazing,” concluded Dhok. 

What are you waiting for? Click here to apply for data science jobs at Yulu today.