The Replica team is creating a next-generation urban planning tool that can help cities answer key transportation and infrastructure questions. Replica allows city planners to understand the full impact of their decisions within the rapidly changing world of mobility.
Replica, previously known as the Model Lab, is a part of Sidewalk Labs, an Alphabet company tackling the challenges of urban growth.
What is the Role?
We’re looking for a Machine Learning Engineer or Data Scientist with experience delivering algorithms and implementing end-to-end data processing pipelines. You will be a part of the team responsible for developing production-grade machine learning models that capture behavioral choices of city dwellers in a privacy-preserving way. You will be training artificial agents to respond to changing travel conditions, new urban infrastructure, and updated public policies in city-wide simulations.
What you have to achieve:
- Within your first month, you will build a deep understanding of Replica’s data protection and location privacy principles, technology infrastructure, coding style and standards, and have contributed several features to the data science code base.
- By the end of month 3, you will be leading the development of a range of algorithms, enforcing a high-quality standard of predictive performance while preserving privacy.
- From there your role will be continue to evolve—we’re a fast growing team. Here’s a small sampling of projects we may work on:
- Improve the robustness and accuracy of data inferences from multiple sources.
- Build and test an end-to-end data processing pipeline for production quality models.
- Explore novel algorithms for learning from spatial/temporal data.
- Know your computer science. Candidates need to be able to hit the ground running. Therefore, 3+ years of experience in a production data science environment is strongly recommended.
- Know your data science. Have a strong hands-on knowledge of machine learning algorithms, both deep and shallow. Have a strong interest or prior experience in reinforcement learning, multi-agent systems and modern AI.
- Bias towards action and shipping. Once you’ve sketched out an idea, you find the fastest path to a prototype to prove the concept. Be proficient in applied optimization: challenges of fast changing cities require solving myriads of new types of optimization problems.
- Be creative: you’ll come up with new ideas based on your broad understanding of technological possibilities and city domain knowledge.
- Have strong collaboration and communication skills.
- Care about cities, solving hard problems, and our team’s success.
- Familiarity with transportation systems and transportation modeling is a bonus.