I am currently building an autonomous marketing mix modelling platform paired with agentic media buyers and would really value input from this community, I already have onboarded one of your community to lead the data science aspect, so I am particularly looking for an full stack engineer with experience in probabilistic modelling and production ML systems.
The goal is to move beyond static MMM towards a continuous, autonomous decision engine that:
-
Continuously ingests spend and performance data
-
Runs automated Bayesian MMM with hierarchical and time varying effects
-
Quantifies uncertainty and marginal returns
-
Uses agents to recommend and execute budget allocation decisions
-
Learns continuously as new data arrives
In short, a closed loop system for measurement, decision making, and execution.
Where I am today
-
Early stage startup
-
Data scientist already onboard with Bayesian experience
-
Initial design partner conversations underway
-
Working through modelling and system architecture
I am now looking for someone to help build the platform and agent layer. This includes:
-
Productionising Bayesian MMM
-
Building continuous training and inference pipelines
-
Designing decision systems using posterior outputs
-
Building agent workflows for budget allocation and optimisation
-
Scaling the platform for multiple brands
Who I am looking for
Someone interested in:
-
PyMC and probabilistic modelling
-
Bayesian inference in production
-
ML systems and platform engineering
-
Decision systems, reinforcement learning, or bandits
-
Early stage startup building
If this sounds interesting, I would really like to connect. Happy to share more context privately.
Also very open to feedback from the community on architecture, modelling approaches, or similar problems others have worked on.
Thanks in advance.