Data Science (Machine Learning for Maps)

The Team

Grab — the leading super app in Southeast Asia — combines transport, food delivery, logistics, payments and much more in a single platform. In all this, our map and localisation infrastructures serve as fundamental building blocks in enabling multiple location-based services.

The world in Southeast Asia constantly changes — and therefore, so must our representation of this world. This not only provides a huge challenge, but also an exciting opportunity to leverage high-volume, high-velocity datasets in various modalities to build complex models that help Grab understand this world intelligently.

The Data Science (Machine Learning for Maps) team at Grab focusses on building map-based intelligence such as localization, routing, travel time estimation, traffic forecasting, that assist various Grab services like transportation, logistics, and pricing. We extensively use Computer Vision, NLP, Information Retrieval and Text Mining along with conventional machine learning methods on a variety of signals including images, videos, text, sensor readings, GPS probes etc to understand our locations and road networks. We also support the development of innovative, highly scalable, models through deep research and advanced analysis so that we make our products intelligent and delight our customers. We foster a culture where we enjoy raising the bar constantly for ourselves and others, and that strongly supports the freedom to explore and innovate.

Duties and Responsibilities

• Understand business needs, identify areas for investigation, translate them to technical problems to be solved

• Build, quick machine learning prototypes (e.g. predictive, forecasting, clustering) using both proven and experimental techniques

• Define hypotheses, develop and execute necessary tests, experiments, and analyses to prove or disprove them

• Translate data speak to human speak by effectively conceptualizing analysis to team members and business stakeholders

Requirements

• Ph.D. graduate, or Masters in Computer Science, Electrical/Computer Engineering, Operations Research or Mathematics/Statistics

• Understanding of machine learning, deep learning, data mining, algorithmic foundations of optimization, probability and statistics.

• Proficient in one or more of the following programming languages: Python, R, Scala, and (or) Golang.

• Desired skills - Sound knowledge of ML/DNN algorithms, hands-on experience using Keras, TensorFlow, or Caffe, as well as Python, Numpy, SciPy, Scikit Learn and Spark ML/MLlib.

• Familiar with relational databases and hands-on experience with SQL

• Self-motivated, independent learner, and willing to share knowledge with team members

• Detail-oriented and efficient time manager in a dynamic working environment

• Able to communicate well in English both verbally and in written communication, as well as convey data insights and results with effective visualisations.

Nice-to-Haves

• 3+ years of industry experience in a Computer Science, Engineering, Operations Research or Mathematics/Statistics capacity

• Experience in curating, manipulating and analysing large geographical or spatio-temporal datasets is a very desirable plus.

• Expertise in one or more specialized domains including Information retrieval, Computer vision, Natural Language Processing, Text mining, and Geo-spatial data mining.

• Experience in processing data and building ML models at scale, using distributed computing frameworks such as Spark/MapReduce.

• Experience in production software engineering (web development, test-driven development, code versioning with Git, code reviews, CI/CD) is a big plus.

• Experience in designing, deploying or maintaining ML systems in production is a big plus.

• Familiarity with modern data pipeline and warehousing stacks, e.g. Hive, Livy, Azkaban/Airflow, PrestoDB, Redshift, Kafka stream processing etc.

• Familiarity with NoSQL frameworks or search engine / indexing frameworks like ElasticSearch

• Familiarity with GIS frameworks and/or databases (e.g. PostGIS).

• Familiarity with cloud platforms like AWS or Azure