Senior ML Engineer / Bayesian Attribution Lead (iGaming)
Immensity.ai
About Immensity.ai
Immensity.ai builds production-grade Bayesian attribution and forecasting systems for iGaming operators. We focus on estimating incremental impact and uncertainty across paid media, affiliates, and lifecycle channels in environments where last-touch attribution and deterministic models fail.
Our modeling approach is intentionally pragmatic: Bayesian regression–based MMM with explicit uncertainty, designed to operate under noisy data, delayed conversions, and shifting spend strategies. Models are used directly to inform budget allocation, marginal ROI estimation, and forward-looking performance scenarios.
Role Summary
We are seeking a senior Bayesian practitioner to lead the development and productionization of our attribution models. This role combines probabilistic modeling, ML engineering, and applied decision-making.
You will own Bayesian models that estimate channel-level incrementality and uncertainty, and help evolve them into reliable, repeatable inference workflows used by operators on a recurring basis. This is a hands-on role focused on models that ship.
What You’ll Work On
Bayesian Modeling & Inference
-
Build and maintain Bayesian regression models for marketing attribution
-
Specify priors that stabilize inference under collinearity, sparsity, and noisy observations
-
Model lagged effects (e.g., adstock), diminishing returns, and nonlinear spend response
-
Quantify uncertainty via posterior distributions and propagate it into forecasts
-
Perform posterior predictive checks and ongoing model validation
PyMC & Probabilistic Tooling
-
Implement and maintain PyMC models used in recurring inference pipelines
-
Tune sampling strategies and diagnose convergence issues
-
Balance model complexity with runtime constraints in production
-
Version models and manage evolution as data and requirements change
ML Engineering & Infrastructure
-
Productionize Bayesian models on AWS (EC2, S3, SageMaker, Lambda)
-
Build training and inference workflows that support frequent re-estimation
-
Integrate model outputs into APIs, dashboards, and decision systems
-
Ensure reproducibility, monitoring, and graceful failure handling
Applied Attribution & Decision Support
-
Translate posterior outputs into actionable attribution and budget insights
-
Work with product and client-facing teams to ensure outputs are interpretable and useful
-
Help define how uncertainty should inform spend decisions rather than be ignored
What We’re Looking For
Required
-
Strong foundation in Bayesian statistics and probabilistic modeling
-
Hands-on experience with PyMC (or Stan / NumPyro with willingness to work primarily in PyMC)
-
Experience building Bayesian regression models for real-world data
-
Strong Python skills and experience deploying models into production
-
Comfort reasoning about priors, identifiability, convergence, and uncertainty
Preferred
-
Experience with marketing mix modeling or attribution
-
Familiarity with adstock, saturation curves, or nonlinear response modeling
-
Experience deploying probabilistic models in cloud environments
-
Exposure to iGaming or similarly noisy acquisition environments
Why This Role
You’ll be working on Bayesian models that directly influence how large marketing budgets are allocated — with real uncertainty, real constraints, and real consequences. This is a role for someone who wants Bayesian inference to matter operationally, not just theoretically.
Please contact Tim Mathews at tim@immensity.ai if you think you could be a good fit for this role!