My Interview on the Supply Chain Optimizers Podcast
Diego Solorzano was kind enough to interview me on his new podcast, Supply Chain Optimizers1
You can find it on YouTube and Spotify.
Diego was a fun interviewer. We covered everything from my business experience to my five hot takes from my supply chain class.
Diego did a nice summary of the episode on LinkedIn. Here is the full version of what he said:
I had the pleasure of chatting with prof. Michael Watson for our Supply Chain Optimizers Podcast. Here's "Everything you wanted to know about AI vs. Optimization but were afraid to ask... "
Mike has an outstanding career in Supply Chain Optimization: a PhD in Industrial Engineering, Founded Opex Analytics (and sold it to Llamasoft), wrote one of the most influential books on Network Design, currently teaches at Northwestern, and has one of the best Substacks on supply chains you can read.
Calling him an eminence falls short of all he has accomplished in this space.
Here's what I learned from him:
1. Model, model, model
- If you want your organization to get good at optimization, you have to get good at modeling.
- This is a skill you can train for; and probably the most important one to develop if you want to do true supply chain optimization.
2. AI, Optimization, Data Science?
- These are all different career paths and skills. Some are in vogue at different moments.
- The magic happens when you combine them together.
- We are very much in an "AI-era", but that looses sight of the fact that using optimization algorithms together with AI models can (and does) yield much better results for companies.
3. AI + Optimization yield true decision making models and tools.
Here's a specific example. You have to figure out the price at which to sell a commodity:
- ML approach: what price should you sell something at.
- Optimization approach : what restrictions are there? for instance, seasonality, capacity and transportation constraints. Thus, what is the price, during that season that minimizes my transportation and capacity costs.
- Thus, ML + Optimization: a true decision-making tool to answer the question more granularly.
4. In business, you always need "a translator". Someone that can bridge the technical concepts and the value for the business.
- The power of true story telling is at an all time high.
- You can become great at story telling by practicing this skill. A tip: Start writing.
- Organizations today can only succeed if they can get their stakeholders to adopt projects
As much progress as AI is making, optimization breakthroughs are also on the horizon with new linear programs making use of GPUs.
Lasty, a tip to find optimization opportunities: think of any tradeoff your company makes. There's an opportunity for a heuristic or optimization model.
Highly recommend listening to this one. It clears a lot of questions on ai vs. optimization and how to use them at your company. Links to the episode in the comments below..
Thanks for the interview Michael Watson. Really enjoyed it!