For anyone following this blog since its first post in May 2007, this will sound like a “broken record” – the key to successful digital transformation is to use an agent-based versus equation-based model.
Here are excerpts from two recent posts:
To better understand the use of generative AI in procurement, you must first understand agent-based modeling and strand commonality – Procurement Insights (Dec. 18th, 2023)
Strand commonality is based on the theory that seemingly unrelated strands or streams of data events have related attributes.
By recognizing that these “attributes” exist and are related, you can link them to achieve a collective optimal outcome using advanced self-learning algorithms. In a production environment with multiple shifting variables – both historical and real-time, the optimal outcome was achieved 97.3% of the time.”
“In 1998, I developed one of the industry’s first algorithm-driven web-based RA solutions with the generous support of the government’s Scientific Research & Experimental Development Program (SR&ED) for the DND.
The problem wasn’t the tech itself but the misalignment of its implementation and spend application. One day, I will have to tell you about the early AI algorithms that drove the engine. However, here is the result – https://bit.ly/3oe5Vql
Here are the critical elements for a successful RA implementation:
– Use an agent-based versus equation-based model
– Understand the importance of “strand commonality,” e.g., tie-in to other depts such as services and external partners such as UPS*
– Understand commodity characteristics, e.g., Dynamic Flux versus Historic Flatline
– Ensure that your algorithms are self-learning, taking into account both historical performance and current market conditions using a weighted scoring model
– Accept that depending on the size of your procurement department, “FTE resource allocation” will happen.
– Introduce a Products Compression Model as the efficiencies you gain on the acquisition side of the process will enable you to lower inventory levels substantially – in one instance, we are able to reduce warehouse inventory requirement by 80 to 90 percent.
* strand commonality also includes factors such as the time-of-day (TOD) orders are placed impact on commodity types.”
What Does This Have To Do With Orchestration?
In Dr. Elousie Epstein’s recent post, she talks about why she is “Bullish about Process Orchestration” – https://bit.ly/3GXTjJN
So, what is orchestration?
“Business process orchestration, or process orchestration, is a management approach that connects your workflows, technology, and people in end-to-end automated processes. At its core, process orchestration means automating and monitoring your workflows to ensure all parts of your process are running smoothly.” – Appian Blog
Agent-based Modeling
What is agent-based modeling used for?
“Agent-based models are computer simulations used to study the interactions between people, things, places, and time. They are stochastic models built from the bottom up, meaning individual agents (often people in epidemiology) are assigned certain attributes.” – Columbia University
“Agent-based modeling and simulation (ABMS) is a relatively new approach to modeling complex systems composed of interacting, autonomous’ agents’. Agents have behaviors, often described by simple rules, and interactions with other agents, which in turn influence their behaviors.” – The Operational Research Society
What’s The Difference?
The real difference is in the “technology first” risk.
Like agent-based modeling, orchestration involves technology. However, the risk with orchestration is that we view it as a technology play versus a people, process, technology model.
Check out my post’s case study: Are you chasing solutions or solving problems? (Part 1 of 3). The best tech and process mapping would only have made a difference if we understood the unique operating attributes of internal departments and external partners – including UPS.
In short, Generative AI and low-code, no-code orchestration tools are not, by themselves, the answer. Success is ultimately driven by leading with stakeholder communication, collaboration, and effective alignment.
Suppose we fall into the “wow, technology is cool” trap and abdicate our leadership responsibility to the tech providers alone**. In that case, we will find ourselves in the same position we were in in 2019 five to ten years from now. Deloitte CPO Survey: Is the digital transformation of procurement a bust?
Key Takeaway: Lead with a PEOPLE-PROCESS-TECHNOLOGY approach.
Procurement Insights
Does the Dr. Epstein Post represent the long-awaited hope that we have finally moved away from equation-based modeling?
Posted on December 26, 2023
0
For anyone following this blog since its first post in May 2007, this will sound like a “broken record” – the key to successful digital transformation is to use an agent-based versus equation-based model.
Here are excerpts from two recent posts:
To better understand the use of generative AI in procurement, you must first understand agent-based modeling and strand commonality – Procurement Insights (Dec. 18th, 2023)
“At the time, the government’s Scientific Research & Experimental Development program funded my research to develop a web-based self-teaching algorithm platform for the Department of National Defence. – https://procureinsights.com/2023/10/02/are-reverse-auctions-good-or-bad-the-debate-continues/
Strand commonality is based on the theory that seemingly unrelated strands or streams of data events have related attributes.
By recognizing that these “attributes” exist and are related, you can link them to achieve a collective optimal outcome using advanced self-learning algorithms. In a production environment with multiple shifting variables – both historical and real-time, the optimal outcome was achieved 97.3% of the time.”
Are Reverse Auctions Good or Bad? The Debate Continues – (Procurement Insights – Oct. 2023)
“In 1998, I developed one of the industry’s first algorithm-driven web-based RA solutions with the generous support of the government’s Scientific Research & Experimental Development Program (SR&ED) for the DND.
The problem wasn’t the tech itself but the misalignment of its implementation and spend application. One day, I will have to tell you about the early AI algorithms that drove the engine. However, here is the result – https://bit.ly/3oe5Vql
Here are the critical elements for a successful RA implementation:
– Use an agent-based versus equation-based model
– Understand the importance of “strand commonality,” e.g., tie-in to other depts such as services and external partners such as UPS*
– Understand commodity characteristics, e.g., Dynamic Flux versus Historic Flatline
– Ensure that your algorithms are self-learning, taking into account both historical performance and current market conditions using a weighted scoring model
– Accept that depending on the size of your procurement department, “FTE resource allocation” will happen.
– Introduce a Products Compression Model as the efficiencies you gain on the acquisition side of the process will enable you to lower inventory levels substantially – in one instance, we are able to reduce warehouse inventory requirement by 80 to 90 percent.
* strand commonality also includes factors such as the time-of-day (TOD) orders are placed impact on commodity types.”
What Does This Have To Do With Orchestration?
In Dr. Elousie Epstein’s recent post, she talks about why she is “Bullish about Process Orchestration” – https://bit.ly/3GXTjJN
So, what is orchestration?
“Business process orchestration, or process orchestration, is a management approach that connects your workflows, technology, and people in end-to-end automated processes. At its core, process orchestration means automating and monitoring your workflows to ensure all parts of your process are running smoothly.” – Appian Blog
Agent-based Modeling
What is agent-based modeling used for?
“Agent-based models are computer simulations used to study the interactions between people, things, places, and time. They are stochastic models built from the bottom up, meaning individual agents (often people in epidemiology) are assigned certain attributes.” – Columbia University
“Agent-based modeling and simulation (ABMS) is a relatively new approach to modeling complex systems composed of interacting, autonomous’ agents’. Agents have behaviors, often described by simple rules, and interactions with other agents, which in turn influence their behaviors.” – The Operational Research Society
What’s The Difference?
The real difference is in the “technology first” risk.
Like agent-based modeling, orchestration involves technology. However, the risk with orchestration is that we view it as a technology play versus a people, process, technology model.
Check out my post’s case study: Are you chasing solutions or solving problems? (Part 1 of 3). The best tech and process mapping would only have made a difference if we understood the unique operating attributes of internal departments and external partners – including UPS.
In short, Generative AI and low-code, no-code orchestration tools are not, by themselves, the answer. Success is ultimately driven by leading with stakeholder communication, collaboration, and effective alignment.
Suppose we fall into the “wow, technology is cool” trap and abdicate our leadership responsibility to the tech providers alone**. In that case, we will find ourselves in the same position we were in in 2019 five to ten years from now. Deloitte CPO Survey: Is the digital transformation of procurement a bust?
Key Takeaway: Lead with a PEOPLE-PROCESS-TECHNOLOGY approach.
** Why it’s time for service provider leadership to step out from behind their organizations’ logos
30
Share this:
Related