Data Scientist

Data Scientists at OLX have a unique platform to work with. Through a “buy and build” strategy, OLX has spread to 40 countries. Focusing on emerging markets, the differences in culture, technological advancement and internet savviness between the countries are huge. OLX is a global operating company that aims to accommodate the needs of all its 300 million user worldwide, as heterogeneous as they may be. Understanding and predicting user behavior correctly is the key to success. After all, online classifieds platforms live of user engagement. They don’t manufacture or sell anything, they connect buyers and sellers. So, the people who can help OLX engage their users are crucial to the business – and those are the Data Scientists.

The vast data stacks of OLX

Through organic growth and acquisition, OLX has become the world’s leading classifieds platform, where consumers sell their second hand goods to other consumers. Or offer services. Or put their house for sale. State of the art in-house technology offers near real-time tracking of all users’ browsing and transactional behavior. Subsequently, the data volume is significant – think hundreds of terabytes. And data is there to be mined for insights. Actionable insights, preferably, that can help create the most relevant, most personalized and most user-friendly experience possible.

This is a virtual greenfield to work on, which touches the core of the business. A new team  of 10 people will be formed around “Search & Relevancy”. The key role in that team is for two Data Scientists. They will apply machine learning to the behavioral data of 300 million worldwide OLX users and come up with the algorithms that help dazzle every single user with a very tailored experience while using OLX. Needless to say, only the best will be able to pull this of.

When Peter goes on the OLX app (or website), the feed he sees should be for him. And for him only. Even before Peter starts his search for that drum kit or inflatable pool, OLX knows what Peter likes to see. Because they have his browsing data, their ad statistics, they ask him questions, they know his Facebook profile, and they have a lot of context. Through machine learning, the Data Scientist can predict what Peter would like to see on his opening feed. So when Peter starts searching, OLX gives him results that are relevant, because the seller is close to his home, or because that seller is always responding quickly during that time of day or because the seller publishes other items they know Peter likes. Or because of other factors. And the Data Scientist works on finding out what those other factors could be.

About the vacancy: Data Scientist

So, what does a Data Scientist do at OLX? They start out with formulating the right questions. The next step is exploring billions of records, developing predictive models, optimizing algorithms and performing deep-dive analysis to help answer these questions. Noteworthy: a 10% algorithm improvement can mean instant 10% incremental net revenues in this business. Data Scientists are responsible for identifying key prediction/classification problems and devising innovative solutions. They assist with product direction by bringing in innovative ideas generated through data mining and data modeling, and they formulate smart recommendations based on performed analysis.

Read the full candidate pack on Big Data at OLX

Personality

• Keen aptitude for large-scale data analysis with a passion for identifying key insights from data and understanding how to convert a data-mining algorithm into a real-time decision engine
• Understanding of the Online or Classifieds/E-commerce industry is a plus
• Cooperative, collaborative and flexible mindset
• Critical thinking

Experience

• Advanced degree in Mathematics, Computational Statistics, Computer Science, Physics or related fields
• Data Science, Predictive Modeling and Programming
• Hands-on with Machine Learning techniques such as logistic regression, naïve bayes, SVM, decision trees, neural networks and random forests
• Structured programming languages (Python – preferentially, R, SAS, etc.)
• Relational and non-relational databasing (SQL, NoSQL, Hadoop etc.) and Data Warehousing
• Fluent in English