Staff Software Engineer, Ads Experimentation Platform

Reddit has a flexible first workforce! if you happen to live close to one of our physical office locations our doors are open for you to come into the office as often as you'd like. Don't live near one of our offices? No worries: You can apply to work remotely from the United States or Canada.

As a company, Reddit primarily generates revenue through advertising, and we're working towards building a massive business to fund our mission. We distinguish ourselves from other digital ad platforms by attracting advertisers who want to connect with a specific target audience because of our passionate and engaged communities.

Reddit Ads marketplace is a dynamic and complicated environment where millions of engagements among Reddit users, advertisers and the platform occur every second. It’s critical for us to extract useful and statistically significant signals from the massive volume of events to guide the product developments and decisions in a scientific way. 

The Ads Experimentation Platform (AEXP) team builds tools for our engineers and scientists to accurately test new products and efficiently verify new ideas in the ads marketplace. Have you heard about Budget Cannibalization? Have you wondered why a user-split A/B testing is insufficient in testing products in the Ads marketplace? How can we make the most out of rare and sparse samples (e.g. conversion events)? How can I explore 100 algorithmic candidates at the same time? These are the typical problems we are dealing with on a daily basis.

Meanwhile, Ads Pacing plays a crucial role in both supporting the dynamism of the marketplace and accommodating the budget-aware A/B experiments, therefore our team is also committed to building scalable and flexible pacing infrastructures that improves the market efficiency and budget-aware testing capacity.

Some of the project highlights include:

• An A/B testing framework that split both users and ad budgets to create parallel and mutually exclusive control/treatment sub ads marketplaces,
• An exploration tool that can run multiple (hundreds of) A/B testing pairs of campaigns over multiple algorithmic candidates for pacing,
• Partitioning the pacing and budgeting systems to accommodate the growing number of advertisers and maintain the refreshing rate above once per minute,

As a Staff Software Engineer in the Ads Experimentation Platform team, you will work closely with data science partners and product leaders to improve the capacity and efficiency of existing experimentation tools, research and build new testing methodologies and tools, and lead the experimentation practices across the Ads marketplace teams. You will be responsible for the quality and technical approach within the team; partner with other leads in direction setting, planning, and overseeing engineering designs and executions; establish and contribute to the group’s culture and processes. You will collaborate with cross-functional engineering, product, and data science teams to ensure that our customers’ voices are represented in your work..

Responsibilities:

• Building Reddit-scale optimizations to improve advertiser outcomes using cutting-edge techniques in the industry.
• Leverage live auction data and model predictions to adjust campaign bids in real time.
• Incorporate knowledge of the Reddit ads marketplace into budget pacing algorithms powered by control & reinforcement learning systems 
• Lead the team on designing new bid & budget optimization products and algorithms as well as conducting rigorous A/B experiments to evaluate the business impact.
• Actively participate and work with other leads to set the long term direction for the team, plan and oversee engineering designs and project execution.

Who You Might Be:

• Degree in a quantitative discipline: engineering, statistics, operations research, computer science, informatics, applied mathematics, economics, etc.
• 7+ years of contributing high-quality code to production systems that operate at scale.
• 5+ years of experience building ads-serving related systems, including but not limited to ads targeting, ads ranking, ads pacing.
• Experience building A/B testing frameworks for multiparty marketplace scenarios, e.g. food delivery service, ride sharing, ads marketplace.
• Experience leading large engineering teams and collaborating with cross-functional partners is required, especially with data science partners.

Required Qualifications:

• Significant experience in one or more backend programming languages, experience with Go or Scala is preferable.
• Significant experience with API development and service frameworks, e.g. Thrift and Protobuf.
• Data processing frameworks like Spark, Flink, Kafka, Druid, etc.
• A cloud service provider like AWS or GCP
• Tools: Kubernetes, Drone, CircleCI, Spinnaker, Argo, Airflow, Docker.
• Datastores: ElasticSearch / Amazon OpenSearch, Redis, Postgres, Cassandra, BigQuery

Benefits:

• Comprehensive Healthcare Benefits
• 401k Matching
• Workspace benefits for your home office
• Personal & Professional development funds
• Family Planning Support
• Flexible Vacation (please use them!) & Reddit Global Wellness Days
• 4+ months paid Parental Leave
• Paid Volunteer time off

#LI-CJ1 LI-Remote