Stephany Lapierre, this will read well in a brochure:
“We could have used other technologies to deliver answers back to users but customers liked the conversational and contextual UX and the addition of agentic search capabilities to provide more context on suppliers (as long as we can expose provenance).”
Let’s try this – What is a self-learning algorithm?
Self-learning models are AI models that, once deployed, can be optimised by training them on data that becomes more available over time. This process prevents engineers from having to begin building new AI models from scratch every single time they collect more data. SOURCE – Monolith
However, and this is key, it only works if you start out with an agent-based rather than equation-based model; otherwise, you have the age-old garbage-in, garbage-out problem that even the greatest tech and data scientists can’t fix.
Garbage-in, Garbage-out AI Version 2024
Here is a simplified, high-level example of an agent-based model – https://bit.ly/3FBnFRr
The basis or heart of the platform I developed in the late 1990s and early 2000s was my theory of strand commonality, in which seemingly disparate streams of data have unique attributes that are actually related and contribute across multiple streams to achieve a desired outcome.
You can check out the results in the above link and drill down even further using this link – https://bit.ly/3oe5Vql
Key Takeaway: Today, they use the buzzword orchestration—back in the late 1990s, we called it the Metaprise—which is how you ensure the right inputs to generate the right equations. Refer to my Dangerous Supply Chain Myths (Part 7) post from 2007.
Even Generative AI can’t replicate an agent-based model—people have to do it!
Hints: Time of Day orders come in, policies and practices of other departments, and external stakeholder considerations to start. Generative AI cannot sift through this – it requires human experience, expertise, and active involvement before introducing any technology.
30
Procurement Insights
Here is a more direct path to procurement’s AI success (Late Night Edition)
Posted on August 12, 2024
0
Stephany Lapierre, this will read well in a brochure:
“We could have used other technologies to deliver answers back to users but customers liked the conversational and contextual UX and the addition of agentic search capabilities to provide more context on suppliers (as long as we can expose provenance).”
Let’s try this – What is a self-learning algorithm?
Self-learning models are AI models that, once deployed, can be optimised by training them on data that becomes more available over time. This process prevents engineers from having to begin building new AI models from scratch every single time they collect more data. SOURCE – Monolith
However, and this is key, it only works if you start out with an agent-based rather than equation-based model; otherwise, you have the age-old garbage-in, garbage-out problem that even the greatest tech and data scientists can’t fix.
Garbage-in, Garbage-out AI Version 2024
Here is a simplified, high-level example of an agent-based model – https://bit.ly/3FBnFRr
The basis or heart of the platform I developed in the late 1990s and early 2000s was my theory of strand commonality, in which seemingly disparate streams of data have unique attributes that are actually related and contribute across multiple streams to achieve a desired outcome.
You can check out the results in the above link and drill down even further using this link – https://bit.ly/3oe5Vql
Key Takeaway: Today, they use the buzzword orchestration—back in the late 1990s, we called it the Metaprise—which is how you ensure the right inputs to generate the right equations. Refer to my Dangerous Supply Chain Myths (Part 7) post from 2007.
Even Generative AI can’t replicate an agent-based model—people have to do it!
Hints: Time of Day orders come in, policies and practices of other departments, and external stakeholder considerations to start. Generative AI cannot sift through this – it requires human experience, expertise, and active involvement before introducing any technology.
30
Share this:
Related