Walmart-owned Flipkart is one of the leading e-commerce companies in India. With more than 400 million registered customers, offering over 150 million products across 80+ categories, the company has enabled millions of consumers, sellers, merchants, and small businesses to be part of India’s digital commerce revolution.
Flipkart leverages data science extensively, where it uses a multimodal interface with an augmented reality camera, voice and vernacular interface to handhold customers. The company said that data science is widely applied in developing and using regional language interfaces, voice assistants, and chatbots to help users with discovery and decision-making processes.
Sign up for your weekly dose of what's up in emerging technology.
Particularly for vernacular interfaces, the share of Indian-language users on Flipkart increased from 12 per cent in Q4 2020 to 18 per cent in Q4 2021, with a greater number of customers adopting e-commerce from Tier 3 cities and beyond. Among the eleven Indian language interfaces on Flipkart, Tamil, Hindi, Telugu, and Bengali are the most widely used languages, growing by over 2X over the past year. Malayalam and Bengali are witnessing the fastest growth in terms of adoption among new customers.
Similarly, Flipkart’s Voice Search saw nearly 3 million daily queries this year, with Tier 3+ cities comprising more than half of the overall voice queries.
Besides these, Flipkart also uses data science and AI to drive personalisation to enable a tailored e-commerce experience for millions of online customers across the country. The company has worked extensively towards personalisation over the years. It can apply it throughout the funnel, including personalisation at multiple touchpoints, whether it is the homepage, product page recommendations or customer service experience.
While data science plays a critical role in shaping the Flipkart user journey and transforming merchandising itself, a combination of tech innovations, including high-end computer processing, natural language processing, machine learning and artificial intelligence, is also used to solve for connecting India’s remotest places with seamless delivery.
“We utilise data sciences across selection and supply ecosystem, discovery (search and recommendation), trust and safety, N200 (bringing customers from smaller towns), customer experience, fintech and supply chain, to name a few,” shared Mayur Datar, Chief Data Scientist at Flipkart.
Betting big on vernacular languages
“As democratising e-commerce has been a key pillar of Flipkart’s efforts, providing the same to customers in their own language will elevate it into an emotional experience from a purely functional experience,” said Datar.
He said India is one of the most culturally diverse countries globally, having 22 scheduled languages and over 19,000 languages spoken across the length and breadth of the country. This makes it extremely important to harness the power of native languages from an e-commerce standpoint to bring its benefits to millions of new customers. “Today, roughly half of our customers come from tier 3 and beyond cities. Hence, there is a need to bring a native language experience to onboard the next 500 million customers,” added Datar.
“While Bharat consumers can understand basic English, they feel more comfortable transacting in their local dialects. This also allows the transition of digital interactions in English from being a functional experience to transforming it into an emotional experience with the vernacular languages,” said Datar.
Today, Flipkart has launched a total of 11 Indian languages interfaces, including Hindi, Tamil, Telugu, Bengali, Punjabi, Malayalam, and Assamese, to name a few. “We translated roughly 5.4 million words for each interface by applying a judicious mix of translation and transliteration to offer a native, colloquial experience to customers,” said Datar.
Inside Flipkart’s data science team
“We started with a humble team, and today over the past years, it has grown to a strong 90 member team with experts from key areas such as natural language processing, computer vision analysis, machine learning and more,” said Flipkart’s chief data scientist.
Flipkart’s team consists of data scientists who work across various verticals within the company, led by research directors and domain experts. The team works across supply chain optimisation, discovery and personalisation, building access to the next set of internet users, etc.
As a homegrown e-commerce marketplace, Flipkart believes in bringing talent that can innovate and introduce technological solutions to further strengthen India’s e-commerce journey and bring millions of customers online.
“Over the years, we have developed various India-first solutions by leveraging technologies such as AI, machine learning, IoT, and robotics, which have helped onboard millions of diverse customers from across geographies,” shared Datar, saying that the data science team has played a key role in creating a personalised and seamless experience for their customers.
Currently, Flipkart is on a hiring spree and is looking to hire data science professionals across experience levels and job roles.
Flipkart has about six-seven rounds of the interview process. This includes –
Flipkart has carefully devised its hiring processes to onboard the best talent. From tech screening to hands-on skills, we have designed our processes to have an overall understanding of the candidates’ skill sets. In the exploratory discussion stage, hiring managers interact with the shortlisted candidates to understand their experience, expectations and aspirations for the future.
Technical phone screening
Before the onsite interview, this part involves the candidate discussing with one of their potential peers. This technical screening round primarily focuses on past work experience, role-eligibility, and a few problem-solving questions relevant to the role the candidate is being interviewed for.
“We explore questions about experience and major projects that they have worked on depending on the role they’re being evaluated for.”
To better understand the candidate and the work they have done in the past, they are allowed to present any work of their choice that they have completed in the past. However, the presentation round is only applicable to candidates who have done significant work or have research experience.
In this discussion, candidates are provided with a dataset and an accompanying problem statement to evaluate their ability to understand and solve the problem. The candidate is expected to explain their observations on the data and defend their modelling choice. At the end of this round, candidates will be expected to submit the code.
Depth in data science
This round focuses on the candidate’s ability to translate an unseen fuzzy problem into a concrete problem statement and solution approach. First, the interviewer will give the candidate an abstract business problem related to a real-world challenge faced by Flipkart, followed by a detailed discussion on how they would think through such a problem.
Depth in mathematical modelling
To understand the candidate’s mathematical understanding of an area of their choice, preferably the one discussed in their presentation, the interviewer may present the candidate with a requirement that needs to be changed to a formulation based on changes in the business requirements. Candidates are assessed on their intuitive understanding of the technical domain and their ability to formally present any modifications that may arise during the discussion.
Team and culture fitment
Flipkart is a people-focused organisation, and culture fitment is an important aspect of the process.
Expectations from candidates
“As a people-centric organisation, we remain true to our core values of being ‘Customer First’ while displaying “audacity” and “bias for action”. These are also some of the qualities that we look to cultivate amongst our recruits, and we encourage independent thought and a go-getter attitude,” said Datar.
Apart from having a robust evaluation in place, Flipkart also looks for candidates who imbibe and resonate with the core values that it stands for. “The pandemic has reinforced the importance of exhibiting agility and innovative thinking to solve business problems, especially in the unprecedented situation that does not have a set playbook. Therefore, going forward, adaptiveness and out-of-the-box thinking will be vital parameters that aspirants will be evaluated on,” added Datar.
What can candidates expect from Flipkart?
Being India’s homegrown e-commerce marketplace driven by the best technology, Flipkart believes it is responsible for creating the right opportunities for new joiners to embrace long-term careers in data science, which it facilitates through a host of innovative programmes and initiatives.
“Our large data sciences team plays a critical role in solving relevant problem statements for a growing population embracing e-commerce in new ways. This gives every team member great learning opportunities, and every individual is encouraged to contribute in different ways, powered by a spirit of agility,” explained Datar.
He said Flipkart’s approach offers in-depth learning experiences, an opportunity to collaborate on interesting projects relevant to the dynamic e-commerce ecosystem and various opportunities to engage with business leaders from various business verticals.
“We also have a strong academic collaboration programme that offers some of our data scientists the opportunity to collaborate with faculty and students from premier universities (IITs, IISc, IIMs, etc.) and conduct research together. Such collaborations often lead to publications in top tier academic conferences and journals,” he added.
“At Flipkart, we understand the importance of fostering a culture of innovation and have undertaken various initiatives to make this a reality. We have created a work environment where the spirit of innovation can be embraced by any employee – from a seasoned professional to a new joinee – stretching across teams,” said Datar.
Further, he said that their culture of innovation has a larger ecosystem focus, where they identify the best opportunities that can be extended to people across the country. He believes that this approach dedicated to developing talent not only within the organisation but also in the industry has enabled them to remain a step ahead while giving people an opportunity to work on and solve India-specific needs.
Why join Flipkart?
“At Flipkart, we are committed to building cutting-edge capabilities to elevate the commerce experience for newer and existing users with data science,” said Datar.
If you are looking for an opportunity to create an impact on India’s evolving e-commerce landscape and solve unique problem statements, Flipkart’s data science team can offer you the right platform to solve unique real-world problems statements.
What are you waiting for? Click here to apply!