Skip to content
Artwork for The Decision Intelligence Lab
The Decision Intelligence Lab · April 22 · 36 min

#27 Dr. Tim Varelmann: Primal Solvers, Inventory Agents & the ML-Optimization Stack

Make better business decisions with data and AI—subscribe to The Decision Intelligence Lab Newsletter at ⁠https://decisionintelligencelab.substack.com/⁠. What happens when the optimization rules you learned no longer apply? Dr. Tim Varelmann, Founder of Bluebird Optimization, an expert for mathematical modeling, algorithms and software development, joins Vijay Mehrotra and Michael Watson to unpack the real mechanics of combining machine learning with optimization. Not the textbook version. The practitioner version. Dr. Tim breaks down how ML and optimization actually combine in practice — beyond just demand forecasting. Three integration patterns, the rise of primal solvers, why "start linear" is outdated advice, and a case study where simulation-based inventory optimization saved millions. Plus: maintainable optimization code, Pareto fronts for business stakeholders, and Warren Powell's policy framework. Chapters 0:00 — Preview & Introduction 1:00 — Meet Tim Varelmann 2:50 — ML + optimization: general trends 3:50 — Three ways to combine ML and optimization 6:06 — Solver landscape evolution 9:45 — ML-optimization integration examples 13:35 — Maintainable optimization code principles 16:20 — ML integration challenges with algebraic modeling 17:30 — Downsides: nonlinearity and scaling issues 18:50 — Is "Start linear" advice still valid? 21:35 — Drift case study: inventory optimization 27:29 — Why closed-form inventory formulas fail 29:50 — Engineering the full solution, demand adjustments 32:00 — Future: Warren Powell framework, policy-based optimization 35:15 — Closing Remarks Follow the show Apple: ⁠https://podcasts.apple.com/in/podcast/the-decision-intelligence-lab/id1811085064⁠ Spotify: ⁠https://open.spotify.com/show/0lFoAVKqJHTYSZNpeN61ou?si=0ae973aab0174b3b⁠ Connect with guest Dr. Tim Varelmann: ⁠https://www.linkedin.com/in/timvarel/ Connect with hosts Prof. Vijay Mehrotra (University of San Francisco): ⁠https://www.linkedin.com/in/vijay-mehrotra-ba9498/⁠ Prof. Michael Watson (Northwestern University): ⁠https://www.linkedin.com/in/michael-watson-07600a1⁠ About the podcast The Decision Intelligence Lab podcast delivers real-world insights for data professionals, business leaders, and anyone seeking to leverage data & AI for smarter decision-making & successful business outcomes. For business inquiries, email at ⁠decisionintelligencepodcast@gmail.com⁠

0:00-36:45

transcript

No transcript — this publisher did not publish one.

show notes

Make better business decisions with data and AI—subscribe to The Decision Intelligence Lab Newsletter at ⁠https://decisionintelligencelab.substack.com/⁠.


What happens when the optimization rules you learned no longer apply?


Dr. Tim Varelmann, Founder of Bluebird Optimization, an expert for mathematical modeling, algorithms and software development, joins Vijay Mehrotra and Michael Watson to unpack the real mechanics of combining machine learning with optimization. Not the textbook version. The practitioner version.


Dr. Tim breaks down how ML and optimization actually combine in practice — beyond just demand forecasting. Three integration patterns, the rise of primal solvers, why "start linear" is outdated advice, and a case study where simulation-based inventory optimization saved millions. Plus: maintainable optimization code, Pareto fronts for business stakeholders, and Warren Powell's policy framework.


Chapters


0:00 — Preview & Introduction

1:00 — Meet Tim Varelmann

2:50 — ML + optimization: general trends

3:50 — Three ways to combine ML and optimization

6:06 — Solver landscape evolution

9:45 — ML-optimization integration examples

13:35 — Maintainable optimization code principles

16:20 — ML integration challenges with algebraic modeling

17:30 — Downsides: nonlinearity and scaling issues

18:50 — Is "Start linear" advice still valid?

21:35 — Drift case study: inventory optimization

27:29 — Why closed-form inventory formulas fail

29:50 — Engineering the full solution, demand adjustments

32:00 — Future: Warren Powell framework, policy-based optimization

35:15 — Closing Remarks



Follow the show

Apple: ⁠https://podcasts.apple.com/in/podcast/the-decision-intelligence-lab/id1811085064⁠

Spotify: ⁠https://open.spotify.com/show/0lFoAVKqJHTYSZNpeN61ou?si=0ae973aab0174b3b⁠


Connect with guest

Dr. Tim Varelmann: ⁠https://www.linkedin.com/in/timvarel/


Connect with hosts

Prof. Vijay Mehrotra (University of San Francisco): ⁠https://www.linkedin.com/in/vijay-mehrotra-ba9498/⁠


Prof. Michael Watson (Northwestern University): ⁠https://www.linkedin.com/in/michael-watson-07600a1⁠


About the podcast

The Decision Intelligence Lab podcast delivers real-world insights for data professionals, business leaders, and anyone seeking to leverage data & AI for smarter decision-making & successful business outcomes.

For business inquiries, email at ⁠decisionintelligencepodcast@gmail.com⁠

links7