Principal Data Scientist, Optimization
Lead development of a marketing analytics platform's media plan and budget optimization capabilities, combining hands-on mathematical optimization work with technical team leadership. A player-coach role for someone who wants to write and ship production optimization code while mentoring a small team.
The company is a leader in incrementality-based media measurement and optimization, helping brands plan, test, and optimize billions of dollars in media investments through automated experimentation and media mix modeling. In this role you'll lead development of the platform's media plan and budget optimization capabilities. It's a hands-on role that also carries team leadership: you'll write and ship production code, and you'll lead and mentor a small team of data scientists doing the same work. In practice, you'll help decide how the company allocates billions of dollars in media spend across its clients, using large-scale mathematical optimization to do it.
Key Responsibilities
- Write, ship, and maintain production code for the platform's media plan and budget optimization capabilities. This is a coding role, and you will stay in the codebase.
- Build and solve large-scale optimization problems (linear, mixed-integer, and nonlinear or constrained) that allocate media budget across channels, tactics, geographies, and time.
- Turn media mix model outputs such as response and saturation curves, adstock, and marginal returns into optimization formulations, and trace problems back to the underlying models when they show up.
- Make sure optimization recommendations hold up against incrementality evidence, including geo experiment results where they exist.
- Lead and mentor a small team of data scientists. Set technical standards, review their optimization work and code, and help them grow.
- Work with Product and Engineering to turn optimization work into documented, scalable platform features.
- Assess new optimization methods, solvers, and AI tools that could improve solution quality or the team's productivity.
- Act as the go-to expert on media plan optimization for internal teams, customers, and sales conversations.
What Success Looks Like
- The optimization engine gets more accurate and faster, handles more clients, and improvements reach production.
- You are writing and reviewing code regularly, and the team's work gets better under your leadership.
- Marketers trust the budget plans and allocation recommendations the platform produces, and those recommendations line up with incrementality evidence.
- Product and Engineering can rely on your team to deliver solid, well-documented work.
- You find useful ways to apply AI tools such as generative AI, LLMs, and agent frameworks (for example, LangGraph or CrewAI) to speed up the team's work.
Ideal Experience
- A PhD (preferred) or master's in a quantitative field with a strong optimization or operations research foundation: operations research, applied mathematics, statistics, econometrics, computer science, or engineering.
- At least 8 years of hands-on experience building and shipping media or marketing optimization, preferably in a SaaS or analytics product, for Retail (including D2C), CPG, or B2B clients.
- Media plan optimization (required, hands-on): extensive experience designing and shipping media plan and budget optimization, allocating spend across channels, tactics, geographies, and time under real business constraints, and turning results into decisions marketers can act on.
- Mathematical optimization (required, hands-on): strong practical command of linear, mixed-integer, and nonlinear or constrained optimization. Hands-on work with solvers such as Gurobi or CPLEX (or open-source equivalents) and modeling tools such as AMPL or Pyomo.
- Media mix modeling (extensive experience required): deep familiarity with how media mix models are built, calibrated, and validated, including response and saturation curves, adstock, constrained regression, and Bayesian methods.
- Hands-on coding (required): expert Python for scientific computing and production code, including NumPy, SciPy, pandas, and optimization and ML libraries. Experience with R, C/C++, C#, or MATLAB is a plus.
- Geo experimentation (a plus): experience designing geo holdout and incrementality experiments, and using the results to calibrate media mix models and optimization outputs.
- A track record of turning optimization and marketing science work into shipped products and clear requirements for engineering.
- Experience leading or mentoring data scientists while staying hands-on yourself.
- Strong communication skills: you can read complex model output and explain what it means to both technical and business audiences.
Perks
- 100% Remote
- Competitive Total Rewards and flexible paid time off
- Opportunities to give back through community volunteer programs
- Engaged, diverse, and curious culture
- Award-winning technology powered by an agile, collaborative team
Compensation
Competitive base salary plus equity, commensurate with experience.