[ Tech stack ]
PyMC
The Python probabilistic programming library for modern Bayesian inference.
PyMC describes a statistical model as a graph of random variables, then fits it with MCMC sampling (NUTS, Hamiltonian Monte Carlo) or variational inference on a PyTensor backend that compiles to C, JAX or Numba. Results come out as ArviZ objects for convergence diagnostics, credible intervals and model comparison. It's our choice when we need to quantify uncertainty, encode domain priors, or work with small datasets using hierarchical models.
[ Why PyMC at Dexon ]
What this technology does well, and why we use it.
Typical usage: Bayesian A/B testing, media mix modeling, forecasting with uncertainty, reliability and risk.
- 01
NUTS sampling and variational inference, with JAX and Numba backends.
- 02
Hierarchical models and explicit priors: robust on small datasets.
- 03
ArviZ diagnostics: convergence, credible intervals, model comparison.
- 04
Ecosystem: PyMC-Marketing (media mix, CLV), pymc-bart, Gaussian processes.
[ Complementary technologies ]
The building blocks we often mobilise alongside.
A stack rarely exists alone. Here are the technologies Dexon most often pairs with this one, through pipeline habits, usage similarity or internal mastery. Click on a building block to see its scope.
[ Reassurance ]
- 0+
- custom projects delivered
- 30
- engineers, designers, project managers
- 80 %
- from top French schools
- 24 h
- average reply time
[ They trust us ]
More than 100 French and European companies trust us
[ Press ]
They talk about us.
Nationwide coverage by BFM Business, Le Figaro, Challenges, La Tribune and CNews. An outside reading of our work and our innovations.
