Logo
HomeAboutBlogWorkshops
Contact us
Footer Icon

The Probabilistic AI
Consultancy

Home
About
Blog

info@pymc-labs.com

Subscribe to our newsletter

Stay connected with the latest developments in Probabilistic AI Statistics and AI.

You can unsubscribe at any time. For more details, review ourPrivacy Policypage.

Contact

© 2025 PyMC Labs. All Rights Reserved.

Table Of Contents
No headings found

Introducing the BETA Release of Our MMM Agent - Powered by PyMC-Marketing

July 11, 2025

By PyMC Labs

We're thrilled to open up BETA access to the latest version of our Marketing Mix Modeling (MMM) Agent - a fully AI-driven assistant built on top of PyMC-Marketing that turns what used to be a multi-month modeling effort into an interactive, informative, and insightful workflow in hours.

Whether you're a data scientist, media analyst, or CMO, this new BETA version is designed to get you to ROI-driven decisions faster - and with more confidence.


✅ What’s New (and Fast)#

Our MMM Agent is more than just another tool - it’s your AI co-pilot. Here’s what it can do:

  • Automated Data Wrangling
    Stop wasting time on reconciling TikTok vs. GA4 inconsistencies. Upload your data and let the agent take care of the prep.

  • Smart Bayesian Modeling
    Automatically selects appropriate techniques for carryover, saturation, and trend detection, tailored to your data.

  • Instant “What-If” Scenarios
    Want to double your Facebook budget? Cut TV spend in half? Simulate it instantly and see the impact on ROI.

  • Causal Insights > Correlations
    This isn’t another black-box MMM. Our agent incorporates experiments, interventions, and external data to give causally robust recommendations.


📈 What’s New Since Alpha#

We’ve made major improvements since the Alpha release, based on your feedback:

  • Inline code and charts within the chat interface
  • Persistent visualizations and tables stored in MLflow
  • Improved modeling diagnostics and debugging tools

🔮 What’s Coming Next#

The BETA is just the beginning. We're working on:

  • 🧾 One-click summary deck generation
  • 💸 An agent to optimize budgets given a fitted MMM
  • 🧪 Custom model priors and functions (e.g., adstock curves, saturation functions)
  • 🧠 Ability to define causal DAGs for deeper, structured modeling

🧪 Join the BETA Today#

We're looking for forward-thinking marketing teams, analytics leads, and data scientists to try out our MMM Agent and shape its next phase.

If you've ever been frustrated by slow MMM cycles, black-box models, or unclear ROI attribution — this is your chance to try something better.

👉 Apply for BETA access and get early insights into what’s working in your marketing mix.

mmm-agent-beta@pymc-labs.com