When starting Lokad back in 2008, we thought we had the solution already all worked out. However, since those early days, there have been many twists and turns along the road, and we have had to change the entire strategy of the company on a large number of occasions.
For this special episode, we look back at the Lokad journey so far, and revisit how the ‘Quantitative Supply Chain’ came to be. We talk more about the various errors we made along the way and how they often were, in fact, the key to helping us grow and evolve.
In addition, we cover a number of fundamental areas, such as forecasting with bias, the drawbacks of common enterprise ‘plug and play’ approaches and the generations of deep learning that helped our solution to evolve to where it stands today. And to really mark the occasion and celebrate our 100th episode, we do all of this live from Paris!
Therefore, we accept various tough and on-point questions from viewers, such as what have been Lokad’s biggest defeats and set-backs so far, more detail on Lokad’s experiences with blockchain, the various supply chain challenges the world has faced due to COVID-19 and how Lokad faces scepticism due to its far from traditional approach to forecasting - to name a few.
To conclude, we discuss why it’s so important to share mistakes and how to learn from them. It’s often hard to admit when you’re wrong, but at the end of the day, it’s the best way to progress rather than finding excuses. Mistakes shouldn’t necessarily be seen as something shameful, but as a sign that things are being tried out and tested, to ultimately be bettered. Lokad doesn’t claim to be perfect, and we will endeavour to continue to honestly evaluate our performance and results.
00:28 Joannes, when we started out this channel, did you really think that we could make it to the 100th episode on a topic like supply chain?
02:29 How did the first couple of years of Lokad go?
03:15 In 2011, you came up with the idea of forecasting with bias using quantile forecasts. Why was this something controversial?
07:46 If you look back at the early years, are there any big mistakes that you made?
10:58 What was the biggest breakthrough from a technical perspective?
13:08 What did you learn from the Bitcoin R&D experience?
15:34 Why did you move away from a more classical data science approach?
19:25 What do you see for the future of Lokad?
22:14 Are there any special reasons for the name “Lokad”?
23:27 What would you say has been your biggest defeat, or setback, in the history of Lokad so far?
25:41 Do you think that Coronavirus highlighted a need to transform “traditional” supply chain models?
28:09 How well did the Lokad algorithm work during the Covid disruptions? Did the customers have to go back to the manual approach?
30:58 Do you see any real value added from blockchain applications in supply chain in the near-term, or is it just more crypto hype?
33:27 How much does the project leader and the Supply Chain Scientist’s professionalism and business understanding impact the result /accuracy of the forecasting?
36:22 Can you give your take on the challenges of implementing narrow AI solutions in business forecasting? Has Lokad looked into this? If not, is data quality an issue that is restricting this?
39:07 Do you think that global supply chains can really last? With climate change, does it still make sense to wear a T-shirt produced in Asia when living in Europe ?
42:16 Since the Quantitative Supply Chain approach is quite different from most traditional methods, do you encounter scepticism? What kind of organizational change management challenges do you encounter when you are implementing Lokad?
45:01 In which area can the Lokad approach unlock the most business value where others cannot?
47:25 What was the biggest problem that motivated taking on a huge task like a complete rewrite?
49:17 How well does Lokad work alongside S&OP type solutions?
51:56 Who writes the blog?
52:59 Why do you think it was important to look back at the Lokad journey?