Machine Learning Engineer at Jupiter

The big vision ✨

As a society, we spent the last decade solving delivery and bringing the physical grocery store online. What we missed is the simple fact that food was already online in the form of content in food communities, blogs and forums. In fact, food makes up a large portion of the social media content we consume today.

At Jupiter, we are pioneering food media & content shopping in the US, and in doing so we're building the next version of grocery shopping. In this new version, customers will create, share and shop content directly from food communities where they find affinity. They'll be able to easily plan their meals and shop for ingredients in a single click, as well as share cooking tips with others — a 10x better way to shop that can only exist online.

We use data science and machine learning to automate the repetitive nature of grocery shopping and we’ve created a social shopping experience that is unique to the world of food. Join us on our mission to build magical food experiences that are convenient, healthy & fun!

[Jupiter](http://jupiter.co) was part of Y Combinator's Summer 2019 batch, and secured 2.8M from an outstanding set of investors including NFX, Khosla Ventures, Canaan Ventures, Switch Ventures, and Paul Buchheit.

About your role 💫

Our “autopilot” predictions and recommendations are the bread-and-butter of the Jupiter platform, and are core to personalizing the Jupiter experience. Joining as a founding member of our machine learning engineering (“Autopilot”) team, you’ll be critical in making foundational decisions around our data stack and model infrastructure at Jupiter to supercharge this system.

As a founding Autopilot engineer, engineering is just as important as data science. In other words, not only will you be responsible for model development and evaluation, but you’ll own productionization of those models, workflows, and ETL as well.

Your responsibilities would include

• Making foundational decisions around our data pipeline and model infrastructure from both an engineering (e.g. workflow management, data warehousing) and data science (e.g. model design/evaluation) perspective.
• Building, launching, evaluating, and maintaining models and custom algorithms to generate predictions of what products or recipes customers are ordering week-to-week, and personalized recommendations for what they might like to add.
• Working closely with the product team to outline feature/UX work which maximizes information provided to models.
• Track prediction/recommendation precision and recall closely across customer segments to inform subsequent modeling decisions.
• Shaping our practice as a founding member of the Autopilot Team, bringing your technical leadership, experience, and knowledge of best practices to establish a strong culture around machine learning engineering.

Ownership within the autopilot team:

• You work primarily on the personalization and recommendation feature sets within the autopilot team.
• You will work with the product team to personalize the store experience and improve product recommendations for customers.

We're looking for the best talent out there, so we're offering competitive base salaries with generous early-hire equity.

About you 👩‍🚀

You might be a great fit if you…

• Have 4+ years of experience productionizing ML systems
• Are thrilled about tackling challenging, seemingly-impossible technical problems around data science, machine learning engineering, predictive modeling, or discrete optimization
• Have lots of experience building out and productionizing systems around personalization and recommendation
• Are excited about learning about and designing complex machine learning algorithms attuned to specific problems or use cases
• Are comfortable working with modern web stacks and integrating data pipelines within production systems - we use TypeScript, Kotlin (JVM), gRPC/protobuf, Terraform, among others
• Are just as excited about data exploration in BigQuery and Jupyter notebooks (i.e. data science) as you are about setting up data pipelines with Airflow (i.e. data engineering)
• Are enthusiastic about learning, trying, and diving deeply into new, unfamiliar technologies
• Have taken a customer-facing product from conception to execution, and are excited about doing so again at Jupiter
• Care about writing clean, well-documented code and appreciate static typing
• Enjoy taking initiative while working with a small team in fast-paced environment
• Are excited about reinventing the online grocery experience, starting with working families