pymc_labs's profile picture. The Bayesian Consultancy • Using PyMC (@pymc_devs) to solve your most challenging data science problems • http://pymc-labs.com

PyMC Labs

@pymc_labs

The Bayesian Consultancy • Using PyMC (@pymc_devs) to solve your most challenging data science problems • http://pymc-labs.com

Marketing measurement shifted fast in 2025. Join our #PyMC Labs roundtable as we break down what actually worked across MMM, CLV, forecasting, and #Bayesian models, and what leaders need for 2026. 👉 Register: dub.link/946GB4c #MarketingMeasurement #BayesianModeling

pymc_labs's tweet image. Marketing measurement shifted fast in 2025. 

Join our #PyMC Labs roundtable as we break down what actually worked across MMM, CLV, forecasting, and #Bayesian models, and what leaders need for 2026. 

👉 Register: dub.link/946GB4c 

#MarketingMeasurement #BayesianModeling

𝗣𝘆𝗠𝗖-𝗠𝗮𝗿𝗸𝗲𝘁𝗶𝗻𝗴 vs 𝗚𝗼𝗼𝗴𝗹𝗲 𝗠𝗲𝗿𝗶𝗱𝗶𝗮𝗻, a real-world benchmark for better #MMM decisions. Join PyMC Labs partner Luca Fiaschi at #PyDataBoston 2025 as he shares a rigorous, open-source comparison of both frameworks using realistic, production-scale data!

pymc_labs's tweet image. 𝗣𝘆𝗠𝗖-𝗠𝗮𝗿𝗸𝗲𝘁𝗶𝗻𝗴 vs 𝗚𝗼𝗼𝗴𝗹𝗲 𝗠𝗲𝗿𝗶𝗱𝗶𝗮𝗻, a real-world benchmark for better #MMM decisions.

Join PyMC Labs partner Luca Fiaschi at #PyDataBoston 2025 as he shares a rigorous, open-source comparison of both frameworks using realistic, production-scale data!

Modeling the SAT Math Gap with #PyMC Why do male test takers score ~30 points higher? Ability or selection bias? At @PyDataBoston 2025, @AllenDowney shows how Bayesian modeling untangles confounding to reveal what’s real. #BayesianModeling #DataScience

pymc_labs's tweet image. Modeling the SAT Math Gap with #PyMC

Why do male test takers score ~30 points higher? Ability or selection bias? At @PyDataBoston  2025, @AllenDowney  shows how Bayesian modeling untangles confounding to reveal what’s real.

#BayesianModeling #DataScience

In #CausalPy 0.6.0, one of the biggest upgrades isn’t a model, it’s clearer docs, tutorials, and examples + a unique Bayesian lens on structural causal discovery using variable-selection priors in joint models. 🧩Try it on Github: dub.link/LpEwFil #BayesianModeling

pymc_labs's tweet image. In #CausalPy 0.6.0, one of the biggest upgrades isn’t a model, it’s clearer docs, tutorials, and examples + a unique Bayesian lens on structural causal discovery using variable-selection priors in joint models.

🧩Try it on Github: dub.link/LpEwFil
#BayesianModeling

#CausalPy 0.6.0 drops an enhanced reporting layer: cleaner tables, consistent summaries, and faster access to “so what?” insights. A big step toward business-ready Bayesian outputs. 🧩Available on Github: dub.link/qL9LBNI #CausalPy #BayesianModeling

pymc_labs's tweet image. #CausalPy 0.6.0 drops an enhanced reporting layer:
cleaner tables, consistent summaries, and faster access to “so what?” insights.

A big step toward business-ready Bayesian outputs.
🧩Available on Github: dub.link/qL9LBNI

#CausalPy #BayesianModeling

🚀 #CausalPy 0.6.0 adds Bayesian Propensity Score Enhancements! Flexible spline adjustments, improved joint modeling, and an example notebook make scoring more robust, showing how design thinking complements #Bayesian estimation. 🧩Try it on Github: dub.link/wAWDbxB

pymc_labs's tweet image. 🚀 #CausalPy 0.6.0 adds Bayesian Propensity Score Enhancements! 

Flexible spline adjustments, improved joint modeling, and an example notebook make scoring more robust, showing how design thinking complements #Bayesian estimation. 

🧩Try it on Github: dub.link/wAWDbxB

🚀 #CausalPy 0.6.0 is live! A key highlight: the new Prior class, enabling fully custom #Bayesian priors and advanced setups like spike-and-slab or synthetic control models. More flexibility, more transparency, better causal modeling. 🔗Github: dub.link/HrRBeQY

pymc_labs's tweet image. 🚀 #CausalPy 0.6.0 is live!

A key highlight: the new Prior class, enabling fully custom #Bayesian priors and advanced setups like spike-and-slab or synthetic control models. 

More flexibility, more transparency, better causal modeling.

🔗Github:  dub.link/HrRBeQY

Bill Engels brought Gaussian Processes to life at @PyDataSeattle 2025. From hierarchical models to a baseball performance case study, this #PyMC-powered talk shows how to model uncertainty with confidence. Watch here: dub.link/XJhEhDg

pymc_labs's tweet card. Bill Engels - Actually using GPs in practice with PyMC | PyData...

youtube.com

YouTube

Bill Engels - Actually using GPs in practice with PyMC | PyData...


Join us tomorrow at @PythonIreland PyCon 2025 for a session with Nathaniel Ford on Bayesian workflows with #PyMC Learn iterative refinement, robust SEM modeling, and the craft behind meaningful statistical discovery. #PyCon2025 #PyMC #DataScience

pymc_labs's tweet image. Join us tomorrow at @PythonIreland PyCon 2025 for a session with Nathaniel Ford on Bayesian workflows with #PyMC 

Learn iterative refinement, robust SEM modeling, and the craft behind meaningful statistical discovery. 

#PyCon2025 #PyMC #DataScience

Can an LLM evaluate ad creative? 🎙️ Our Partner @lfiaschi86 joins @eric_seufert on the Mobile Dev Memo podcast to discuss our recent research paper with Colgate-Palmolive on using LLMs & Synthetic consumers to score product concepts. Upcoming Webinar dub.link/deAEgkv


Can AI really replicate human purchase intent? Our viral study with Colgate showed that LLMs + Semantic Similarity Rating (SSR) can match human intent with up to 90% reliability. Join PyMC Labs to explore how Synthetic Consumers are redefining research 👉dub.link/KGyuRL5

pymc_labs's tweet image. Can AI really replicate human purchase intent?

Our viral study with Colgate showed that LLMs + Semantic Similarity Rating (SSR) can match human intent with up to 90% reliability.

Join PyMC Labs to explore how Synthetic Consumers are redefining research
👉dub.link/KGyuRL5

¡Los modelos también hablan español! 🇪🇸 PyMC Marketing ya está disponible en español! And for our non-Spanish speakers 🌍 PyMC Marketing is now live in Spanish! 🧩 Explore: dub.link/yhgL9Q4 #PyMC #OpenSource #MarketingAnalytics

pymc_labs's tweet image. ¡Los modelos también hablan español! 🇪🇸
PyMC Marketing ya está disponible en español!

And for our non-Spanish speakers 🌍 
PyMC Marketing is now live in Spanish!

🧩 Explore: dub.link/yhgL9Q4
 #PyMC #OpenSource #MarketingAnalytics

Who you gonna call when your model’s haunted by unexplained variance? 👻 PyMC Labs. 🎃 Happy Halloween from the PyMC Labs team, Stay spooky, stay Bayesian. #Halloween #DataScience #PyMC


Measuring how brand awareness or reach drives business results is tough. The new notebook in PyMC-Marketing v0.17.0 uses causal reasoning + joint models to connect upper-funnel efforts to real outcomes. 🧩𝗔𝘃𝗮𝗶𝗹𝗮𝗯𝗹𝗲 𝗼𝗻 𝗚𝗶𝘁𝗛𝘂𝗯: dub.link/1D5riw3 #PyMC

pymc_labs's tweet image. Measuring how brand awareness or reach drives business results is tough.

The new notebook in PyMC-Marketing v0.17.0 uses causal reasoning + joint models to connect upper-funnel efforts to real outcomes.

🧩𝗔𝘃𝗮𝗶𝗹𝗮𝗯𝗹𝗲 𝗼𝗻 𝗚𝗶𝘁𝗛𝘂𝗯: dub.link/1D5riw3
 #PyMC

We're excited to highlight Multi-Objective Optimization in PyMC-Marketing v0.17.! A Bayesian way to balance multiple marketing goals, conversions, engagement, budget, and find the best trade-offs. 🧩 𝗔𝘃𝗮𝗶𝗹𝗮𝗯𝗹𝗲 𝗼𝗻 𝗚𝗶𝘁𝗛𝘂𝗯: dub.link/sLUrY2F #PyMC

pymc_labs's tweet image. We're excited to highlight Multi-Objective Optimization in PyMC-Marketing v0.17.!

A Bayesian way to balance multiple marketing goals, conversions, engagement, budget, and find the best trade-offs.

🧩 𝗔𝘃𝗮𝗶𝗹𝗮𝗯𝗹𝗲  𝗼𝗻 𝗚𝗶𝘁𝗛𝘂𝗯: dub.link/sLUrY2F
#PyMC

This week at @_odsc AI West 2025, we're hosting a hands-on #PyTensor tutorial! Learn how the computational graph library powering #PyMC makes probabilistic models scalable & efficient, supports complex inference, and drives real-world AI decisions. #ODSC2025

pymc_labs's tweet image. This week at @_odsc  AI West 2025, we're hosting a hands-on #PyTensor tutorial!

Learn how the computational graph library powering #PyMC makes probabilistic models scalable & efficient, supports complex inference, and drives real-world AI decisions.

#ODSC2025

Is your product innovation truly driving growth, or just shifting sales around? Discover how causal sales analytics with nested logit models helps brands separate incremental growth from cannibalization. Read more 👉 dub.link/LmaeDBD #CausalAnalytics #ConsumerGoods


Most dashboards show correlations, not causation A mistake that can cost millions. Social media may boost awareness → visits → sale Optimize the wrong step & budgets are wasted. AI-powered causal discovery reveals what truly drives results. Read more: dub.link/odjpeko


We're excited to highlight ShiftedBetaGeoModel (sBG) in PyMC-Marketing v0.17.0! Predict customer retention, model cohorts, and make smarter CLV-driven decisions — all with a single Bayesian model 🧩 Check it out: dub.link/EoYryvG #PyMC #CohortAnalysis #CustomerAnalytics

pymc_labs's tweet image. We're excited to highlight ShiftedBetaGeoModel (sBG) in PyMC-Marketing v0.17.0!

Predict customer retention, model cohorts, and make smarter CLV-driven decisions — all with a single Bayesian model

🧩 Check it out: dub.link/EoYryvG
#PyMC #CohortAnalysis #CustomerAnalytics

Customer Lifetime Value (CLV) is critical, but calculating it takes weeks of data wrangling & modeling. Meet the CLV Agent from PyMC Labs: built on Bayesian foundations to automate CLV analysis end-to-end. Alpha access now open — limited spots. 👉Reach out at [email protected]


Loading...

Something went wrong.


Something went wrong.