Senior Data Scientist, Lyft
Algorithms, Lyft Ads
$148-185k
More information about location
Ride sharing company
Be an early applicant
Ride sharing company
1001+ employees
Be an early applicant
$148-185k
More information about location
1001+ employees
Company mission
To improve people’s lives with the world’s best transportation.
Role
Who you are
- Master’s or PhD in Machine Learning, Computer Science, Optimization, Statistics, Engineering, Applied Mathematics, or a related quantitative field; or equivalent high-impact industry experience
- 5+ years of applied science or machine learning experience, with a track record of deploying production models that drive measurable business outcomes
- Demonstrated ability to own multi-project modeling scope across ambiguous problem spaces and integrate work across engineering, product, and data science partners
- Deep expertise in:
- Ranking and relevance modeling
- CTR/CVR prediction, calibration, and uncertainty modeling
- Optimization and pacing algorithms
- Auction dynamics or marketplace delivery systems
- Causal inference methods for ads measurement
- Strong proficiency in Python, ML frameworks (PyTorch, TensorFlow, JAX, scikit-learn), and distributed data systems (Spark, Snowflake, Databricks)
- Proven experience building large-scale, production-ready ML systems, including model servers, training pipelines, monitoring/alerting, and real-time inference services
- Ability to define and execute offline and online evaluation strategies, including experiment design, counterfactual analysis, and diagnostics for model/system failures
- Strong technical leadership skills — able to align partners, influence technical architecture, challenge assumptions, and guide cross-team modeling decisions
- Demonstrated ability to mentor other scientists, elevate technical quality, and improve modeling/analysis standards across the team
- Excellent communication skills, with the ability to articulate complex modeling concepts, system trade-offs, and scientific reasoning to both technical and business stakeholders
- Track record of driving measurable improvements to model performance, advertiser outcomes, or system efficiency through innovative modeling or optimization techniques
What the job involves
- We are seeking an Algorithms Scientist to help build the next generation of ads relevance, targeting, optimization, and measurement algorithms that power the Lyft Ads platform
- In this role, you will work across large-scale datasets and complex real-time systems to design, prototype, and deploy production-grade machine learning models
- You’ll collaborate closely with Engineering, Product, Data Science, and Sales to translate ambiguous business and advertiser needs into rigorous algorithmic solutions that improve ad performance, enhance marketplace efficiency, and drive meaningful revenue growth
- This is a high-impact, highly technical role within a rapidly scaling business line
- The ideal candidate brings strong applied machine learning intuition, hands-on modeling experience, and the ability to write clean, efficient production code
- You will play a critical role in shaping how advertisers connect with Lyft riders—pushing the boundaries of personalization, measurement, and real-time optimization in a dynamic marketplace
- Lead multiple high-impact Machine Learning and AI initiatives across the Lyft Ads platform — including relevance, targeting, bidding, pacing, delivery optimization, conversion prediction, and measurement systems
- Define the modeling strategy, technical roadmap, and success metrics for ML components that power ad-serving and advertiser performance, ensuring alignment with business and revenue goals
- Own complex, open-ended problem spaces, breaking down ambiguous advertiser, marketplace, and system constraints into well-structured modeling approaches and scientific requirements
- Design, develop, and deploy advanced machine learning, optimization, and decisioning algorithms for large-scale real-time and batch systems, balancing scientific rigor with practical engineering constraints (latency, throughput, cost, reliability)
- Partner deeply with Ads Engineering, Infra, and Product to architect production-grade ML systems — including feature stores, training pipelines, online scoring services, monitoring, A/B frameworks, and model governance processes
- Establish robust evaluation frameworks, defining offline metrics, calibration checks, counterfactual methods, experiment designs, and long-term measurement strategies to ensure model correctness and system stability
- Diagnose systemic issues (drift, feedback loops, cold start, pacing imbalance, auction inefficiencies) and lead cross-functional efforts to improve model performance, user experience, and advertiser ROI
- Drive algorithmic innovation by introducing new techniques from ranking, causal inference, reinforcement learning, probabilistic modeling, graph ML, or optimization, and evaluating their feasibility for large-scale ads systems
- Build reusable modeling infrastructure, libraries, and best practices, enabling faster iteration and higher modeling quality across the broader Ads Science and Engineering teams
- Mentor and guide junior/mid-level scientists and MLEs, serving as a technical advisor on modeling design, experimentation, code quality, and scientific reasoning
- Represent Algorithm Science in cross-functional planning, ensuring algorithms are grounded in strong methodology and aligned with Ads business priorities, advertiser needs, and platform constraints
Our take
Lyft has aimed for consistent growth, making sure to pitch itself as a more reliable and friendly ride-sharing option than its main competitor, Uber. Uber has a greater market share and operates globally, but Lyft hopes to compete by its focus on ride-sharing over Uber's more diversified business approach and as a more rider-friendly and greener alternative to Uber.
Operating only in the USA and Canada, Lyft is in a position of being less affected by global events than Uber but also more vulnerable to local conditions. As an example, while it doesn't face the regulatory hurdles around employment rights that Uber does, it is unable to balance out the cost of the unusually expensive US auto insurance across its operations.
In 2022, Lyft acquired PBSC Urban Solutions, a bike-share equipment and technology supplier, allowing Lyft to compete in new verticals and reaffirming its commitment to green transportation. It has also trialed and launched Lyft Assisted, where drivers help passengers from their door into the car to get to medical appointments. These developments chime well with Lyft's unique selling point as a friendlier and more environmentally aware ride-sharing app.
Steph
Company Specialist at Welcome to the Jungle
Insights
-2% employee growth in 12 months
Company
Funding (last 2 of 14 rounds)
Jun 2018
$600m
LATE VC
Mar 2018
$200m
LATE VC
Total funding: $4.9bn
Company benefits
- Great medical, dental, and vision insurance options
- Mental health benefits
- In addition to 12 observed holidays, salaried team members have unlimited paid time off, hourly team members have 15 days paid time off
- 401(k) plan to help save for your future
- 18 weeks of paid parental leave. Biological, adoptive, and foster parents are all eligible
- Pre-tax commuter benefits
- Lyft Pink - Lyft team members get an exclusive opportunity to test new benefits of our Ridership Program
- Family building benefits
- Lyft Pink - Lyft team members get an exclusive opportunity to test new benefits of our Ridership Program
Company HQ
China Basin, San Francisco, CA
Leadership
John Zimmer
(Co-Founder)Originally an Analyst at Lehman Brothers, Zimmer then founded Zimride which was a private ridesharing company that pivoted into Lyft.
Logan Green
(Co-Founder)Founded Zimride, a private ridesharing company which later pivoted into Lyft.
Salary benchmarks
We don't have enough data yet to provide salary benchmarks for this role.
Submit your salary to help other candidates with crowdsourced salary estimates.
Share this job
View 85 more jobs at Lyft