Machine Learning Engineer- I (Monetization AI)

What does the team do?

The Monetization AI team is responsible for driving meaningful interactions between users of Sharechat and advertisers so as to facilitate discovery and drive commerce. These interactions, if personalized properly, drives incremental business for advertisers and helps users discover services/products. We are also responsible to generate revenue for Sharechat so that we can continue delivering hyperlocal trending experiences to the users.

To ensure that this ecosystem of users, advertisers, and Sharechat is delivering value to everyone, we operate at the confluence of machine learning, personalization, behavioral economics, auction theory and mechanism design, testing our hypotheses through large scale A/B testing and making sure we can deliver while ensuring user privacy. We strive to solve large scale regression problems under extreme data sparsity (prediction models for clicks, conversion, transactions etc.), devise efficient auction mechanisms so that we can deliver ROI to advertisers without manual intervention, perform privacy aware learning so that we can strive for minimal user data without compromising on the value generated.

What You’ll Do?

We are looking for passionate and creative data scientists who have demonstrated the ability to hypothesize ML solutions and deliver measurable impact for business. The candidate should be capable of independently formulating the ML solution by identifying the right signals, experiment and identify the appropriate modeling technique, and perform statistical analysis of the A/B testing. The role gives you an opportunity to:

• Identify the gaps in the current ad serving system, define the opportunity size and hypothesize the ML solution to deliver value to users and advertisers.
• Formulate the hypothesis to validate the statistical stability of the proposal, computational efficiency, and robustness under limited data scenarios.
• Code up the solution and conduct extensive offline experiments, and validate the solution through online A/B tests.
• Write detailed experimentation reports to describe the proposal and reasons for its success/failure.
• Collaborate with peer data scientists, engineering and product teams.
• Advance the state of the art in the field of computational advertising and represent Sharechat by publishing research papers and/or patents.

Who are you?

• Experience of 1+ years for MLE-1.
• Proficiency in Python and good knowledge of various deep learning modeling frameworks such as PyTorch, Tensorflow, or Keras.
• Proficiency in machine learning and understanding of broad modeling techniques, with an understanding of deep learning modeling techniques.
• Past experience in building end-to-end ML models including training, experiment design, ablation studies and deployment of solutions.
• Experience with regression and classification problems under missing data scenarios.
• Demonstrable ability to formulate ML solutions for novel problems
• Hands-on experience with visualization toolkits such as matplotlib.
• Awareness of recent research publications (in their current domain of work) and ability to translate them into working code.
• Ability to work in an ambiguous problem setting and collaborate with others to drive clarity. Strong focus towards action and rapid iteration in a fast paced environment.
• Ownership of the problem and drive to deliver results for the customer through the iterative development of deep tech.
• Experience with computational advertising will be a big plus.
• Prior experience working with cloud stack such as Azure ML, GCP, or AWS is a plus.
• Understanding of statistics, combinatorial optimization is a strong plus.
• Exposure to auction theory and mechanism design is a strong plus.