About InMobi
InMobi has built a global Advertising Platform that powers our customers’ growth by helping them engage their audiences and drive real connections. Over the last 17 years, InMobi has built a second unicorn, Glance, which is advancing digital consumption and creating a new wave of disruption. Present on 400M devices across India, SEA, Japan and the US – Glance is one of the largest content platforms globally with ~200M daily active users.
Position Summary
There are trillions of events a day in our system, requiring models to run at a tremendous scale with millisecond latency. Our learning loop is measured in hours and minutes, making it one of the fastest model-learning playgrounds globally. We have built an infrastructure that enables rapid and scalable model deployment. As data scientists, we collaborate with engineering colleagues to deploy our models. With a growing variable set of hundreds of potential features, this is a highly fertile environment for building, experimenting, refining, and achieving real impact from your models. Immediate bottom-line impact is visible when models fire, allowing you to quickly see the value of your work.
The Experience You'll Need
- A core foundation in Mathematics, Statistics, Algorithms, Optimization, and competent ability in coding with data science languages and tools like Python or Apache Spark.
- A passion to investigate and learn from data, ask provocative questions, and be driven to put real models into production that drive business value.
- Basics of big data processing and cloud computing are critical.
- Open to diverse academic backgrounds, with an intent to think and problem-solve like a data scientist.
- Master’s in a quantitative field such as Computer Science, Statistics, Electrical Engineering, Statistics, Mathematics, Operations Research or Economics, Analytics, Data Science. Ph.D. is a huge plus.
- Depending on the level, experience in the Ad Tech Industry working in Data Science teams, applying algorithms and techniques from Machine Learning, Statistics, Time Series, or other domains to solve real-world problems on large datasets.
- Passion for Mathematics, Algorithms, Machine Learning, and eagerness to apply cutting-edge science to InMobi business problems. Excited by the real-world impact of models in production and fast execution.
- Intellectual depth to translate fuzzy business problems into rigorous mathematical problem statements and algorithms. Experience and passion in troubleshooting when ML models don't produce production lift.
- Comfortable with software programming and statistical platforms such as R, Python, and the big data ecosystem. Experience in Apache Spark is a bonus.
- Comfortable collaborating with cross-functional teams.
- Excellent technical and business communication skills, able to present technical ideas simply to business counterparts.
- High degree of curiosity and ability to rapidly learn new subjects and systems.
The Impact You'll Make
- Lead data science efforts for one of the biggest in-app programmatic exchanges globally, involving project ideation, conceptualization, solution design, measurement, iteration, coaching, deployment, and post-deployment management.
- Design, develop, and test product experiments, guiding the team in practical experiments, product design, model development, and evaluation. Agile iteration across experiments to deliver go-to-market ready products is vital.
- Actively analyze data, design and develop models, and problem-solve solutions with the team.
- Manage stakeholders, serving as the interface with internal Product, Engineering, Data, Infrastructure, and Business teams.
- Contribute to thought leadership in the sector by writing blogs, commentary, and case studies, and speaking at industry conferences.
- Learn to design and build models for specific business problems, identify areas where AI can be applied for business impact, and connect model impact with measurable business outcomes, anchoring in business context and end-user needs.
- Work in a multi-functional team environment, collaborating with diverse individuals from engineering, product, business, campaign management, and creative development teams.
- Experiment with multiple algorithms, learning from building, launching, and reviewing performance to understand why something worked or didn't, and how to tailor techniques.
- Become creative in designing successful models that fit particular problems and are modified to perform.