Back in late 2018 and early 2019, before the GENizah Artificial Idiocy craze began, the doctor did a sequence of AI Series (totalling 22 articles) on Spend Matters on AI in X Today, Tomorrow, and The Day After Tomorrow for Procurement, Sourcing, Sourcing Optimization, Supplier Discovery, and Supplier Management. All of which was implemented, about to be implemented, capable of being implemented, and most definitely not doable with, Gen-AI.

To make it abundantly clear that you don’t need Gen-AI for any advanced back-office (fin)tech, and that, in fact, you should never even consider it for advanced tech in these categories (because it cannot reason, cannot guarantee consistency, and confidence on the quality of its outputs can’t even measured), we’re going to talk about all the advanced features enabled by Assisted and Augmented Intelligence that are (or soon will be) in development (now) and you will see in leading best of breed platforms over the next few years.

Unlike prior series, we’re identifying the sound, ML/AI technologies that are, or can, be used to implement the advanced capabilities that are currently emerging, and will soon be found, in Source to Pay technologies that are truly AI-enhanced. (Which, FYI, may not match one-to-one with what the doctor chronicled five years ago because, like time, tech marches on.)

Today we continue with AI-Enhanced Supplier Discovery that is in development “today” (and expected to be in development by now when the first series was penned five years ago) and will soon be a staple in best of breed platforms (and may be found emerging in development beta versions of some platforms). (This article sort of corresponds with AI in Supplier Discovery The Day After Tomorrow that was published in March, 2019 on Spend Matters.)

TOMORROW

Intelligent Supplier Discovery

How is this different than deep capability match that is available today? Because it sure sounds like the ability to match to a very detailed request is “intelligent”. Compared to most search capabilities in most platforms, it is. But there’s more to selecting a supplier, especially one with whom you need a long term relationship, than just tech specs and certification checks. There are also performance considerations, innovation ability (hard to measure), culture, and other, softer factors.

First of all, you need a platform that can predict the ability of a given supplier to innovate and, more importantly, innovate for you based upon your specific needs. To do this, you need to chart the “innovation history” of a supplier (how many innovations per year, typical gap between innovations), compare the “innovation history” to other suppliers in the industry and category, use a predictive curve fitting or other ML algorithm to predict it’s rate (vs. the average). This is a lot of semantic processing to identify innovations and approximate dates, a lot of trend analysis to find the right predictive algorithms, and a lot of calculation. And then you need the ability to refine the innovation rate by category for a multi-category supplier so the trend line matches your need, and not your competitor’s.

Secondly, you need to be able to parse the “reviews” not just for sentiment, but positive or negative interpretations of specific, relevant “soft factors” like communication, working culture, etc. and compute appropriate ratios or bands that can be compared and be considered in super search / match criteria that is relevant to your organization. Next generation targeted sentiment analysis on factors identified on deep semantic analysis. No Gen-AI needed, just domain specific refinements of traditional approaches (trained on highly vetted, validated data sets).

Predictive Smart Search

For a company in direct manufacturing, electronics, pharma, or another industry where advanced innovation at a fairly rapid pace is required not just for growth, but continued market share retention, identifying the right suppliers is critical. This requires a very deep search, and for specific projects, potentially dozens of requirements and validations that need to be done before a supplier can be invited to an event.

So many in fact that, even if a buyer could identify all of these up front, building the search criteria to capture them all could be difficult. Next generation platforms will learn from each search entered into a platform for a product, category, supplier, etc. and extract the typical criteria, the frequency, and the preferences by organization and user.

Based on this data, when a buyer, new product development specialist, etc. starts a search for a new supplier in a category and/or for a product, the platform will predict which factors are relevant to the user, recommend those factors and factors, and intelligently build the right search and tolerances for the user. And then retrieve the best suppliers, ranked with match percentages.

None of this requires Gen-AI. Frequency is just frequency mapping by product, category, and supplier. Matches are matches as per deep search. Auto query creation is rules based automation. Soft factors are identified by semantic and sentiment analysis. And so on. It just requires a lot of Human Intelligence (HI!) to put it all together.

Is That All, Folks?

Probably not. The more data that is collected, the more analysis that can be done, and the more matching and prediction that can be done across people, products, services, and solutions. And the more “intelligence” (which CAN NOT be generated by Gen-AI) that can be put forward beyond your search before you invite a supplier to an event. But it’s the next step, and we’re going to stop here because we are going to refresh our series on Supplier Management as well.

SUMMARY

Now, we realize some of these descriptions are dense, but that’s because our primary goal is to demonstrate that one can use the more advanced ML and AI technologies that already exist, harmonized with corporate, market and community data, to create even smarter Supplier Discovery applications than most people (and last generation suites) realize, without any need (or use) for Gen-AI, that the organization can rely upon to reduce time, tactical data processing, and risk while increasing supplier intelligence and overall organizational performance. It just requires smart vendors who hire very smart people who use their human intelligence (HI!) to full potential to create brilliant Supplier Discovery applications that buyers can rely on with confidence no matter what category or organization size, always knowing that the application will know when a human has to be involved, and why!

Back in late 2018 and early 2019, before the GENizah Artificial Idiocy craze began, the doctor did a sequence of AI Series (totalling 22 articles) on Spend Matters on AI in X Today, Tomorrow, and The Day After Tomorrow for Procurement, Sourcing, Sourcing Optimization, Supplier Discovery, and Supplier Management. All of which was implemented, about to be implemented, capable of being implemented, and most definitely not doable with, Gen-AI.

To make it abundantly clear that you don’t need Gen-AI for any advanced enterprise back-office (fin)tech, and that, in fact, you should never even consider it for advanced tech in these categories (because it cannot reason, cannot guarantee consistency, and confidence on the quality of its outputs can’t even be measured), we’re going to talk about all the advanced features enabled by Assisted and Augmented Intelligence that were (about to be) in development five years ago and are now available in leading best of-breed systems. And we’re continuing with Supplier Discovery.

Unlike prior series, we’re identifying the sound, ML/AI technologies that are, or can, be used to implement the advanced capabilities that are currently found, or will soon be found, in Source to Pay technologies that are truly AI-enhanced. (Which, FYI, may not match one-to-one with what the doctor chronicled five years ago because, like time, tech marches on.)

Today we continue with AI-Enhanced Supplier Discovery that was in development “yesterday” when we wrote our first series five years ago but is now available in mature best of breed platforms for your Supplier Discovery success. (This article sort of corresponds with AI in Supplier Discovery Tomorrow that was published in March, 2019 on Spend Matters.)

TODAY

Deep Capability Match

As noted in our original posting, if you want a custom produced FPGA, an industrial strength power converter that can handle feeds from your wind farms and water wheels, or a new state-of-the-art surround sound system, you don’t want just any supplier. This is especially true if all they do is produce a fixed set of products, use production technology that is not appropriate for the design you want, have a record of sourcing inferior raw materials, or don’t have the right quality processes in place.

So, when we last tackled this subject five years ago, the new/leading supplier discovery platforms were working on deep capability match that could take a set of requirements for a product, or even a bill of materials, and find matching suppliers for the parts.

Especially since all this needed was deep capability identification and tagging across categories, products, and services that included production process, certifications, materials, etc. Which means that deep capability match was essentially just a super smart search capability across not just a few, but dozens of requirements — as long as the data was properly structured and indexed.

This requires the ability to crawl websites and extract all text and documents, OCR those documents to text, and then semantically process for the relevant information along the recorded dimensions. This just required classical semantic processing which uses ontologies, semantic networks, and custom trained (neural) networks for POS/concept identification when classical processing is not sure. Tech that has now been around and ready for production use for over 15 years. The big challenge was the magnitude of data that needed to be processed and indexed, which is not a problem anymore given the processing power of racks, the size of modern data centres (which require 10X to 100X the processing power for the Gen-AI trainwrecks that don’t deliver), and modern distributed processing algorithms and technology.

And, of course the ability to do rapid semantically aware reg-ex (across similar key words / phrases) for anything not indexed, or indexable in a standard taxonomy.

Resource Capability Match

Sometimes you need very specialized services. As we noted five years ago, for new product design, you need an engineering resource who has designed similar products and is familiar with the new production technologies and components that are on the market. For software implementation, you need a team who has installed the current software in a similar environment that has the same ERPs, OSs, data sources, etc. For utility installation, you need engineers with the right skills and certifications. And so on.

This is essentially just a variant of deep capability match, except you are matching on the services capabilities and the individual’s resumes. Getting here was just determining everything that was relevant for a service, processing large amounts of data, tagging and indexing it appropriately, and supporting very deep multi-faceted searches, using the same semantic technology as described above, but tuned for different service (instead of product) domains.

That’s All For Now, Folks!

Again, focus on supplier discovery was, and still is, limited as there were, and still are, only a few vendors doing it. The good news is that we’re starting to see the technology predicted for “tomorrow” five years ago starting to emerge in these platforms as well.

SUMMARY

Now, we realize some of these descriptions, like yesterday’s, are also quite brief, but again, that’s because this is not entirely new tech, as the beginnings have been around for a few years, have been in development and discussed as “the future of” Supplier Discovery tech before Gen-AI hit the scene, and all of these capabilities are pretty straight-forward to understand. And, if you want to dive deeper, the baseline requirements for most of these capabilities were described in depth in the doctor‘s March 2019 articles on Spend Matters. The primary purpose of this article, as with the last, was to explain how more sophisticated versions of traditional ML and AI methodologies could be implemented in unison with human intelligence (HI!) to create smarter Supplier Discovery applications that buyers could rely on with confidence.

Back in late 2018 and early 2019, before the GENizah Artificial Idiocy craze began, the doctor did a sequence of AI Series (totalling 22 articles) on Spend Matters on AI in X Today, Tomorrow, and The Day After Tomorrow for Procurement, Sourcing, Sourcing Optimization, Supplier Discovery, and Supplier Management. All of which was implemented, about to be implemented, capable of being implemented, and most definitely not doable with, Gen-AI.

To make it abundantly clear that you don’t need Gen-AI for any advanced enterprise back-office (fin)tech application, and that, in fact, you should never even consider it for advanced tech in these categories (because it cannot reason, cannot guarantee consistency, and confidence on the quality of its outputs can’t even be measured), we’re going to talk about all the advanced features enabled by Assisted and Augmented Intelligence (as we don’t really have true appercipient [cognitive] intelligence or autonomous intelligence, and we’d need at least autonomous intelligence to really call a system artificially intelligent — the doctor described the levels in a 2020 Spend Matters article on how Artificial intelligence levels show AI is not created equal. Do you know what the vendor is selling?) that have been available for years (if you looked for, and found, the right best-of-breed systems [many of which are the hidden gems in the Mega Map]). And we’re going to continue with Supplier Discovery. (Find our series on Advanced Procurement — No Gen-AI Needed! Yesterday, Today, and Tomorrow and our series on Advanced Sourcing — No Gen-AI Needed! Yesterday, Today, and Tomorrow through the embedded links.)

Unlike prior series, we’re going to mention some of the traditional, sound, ML/AI technologies that are, or can, be used to implement the advanced capabilities that are currently found, or will soon be found, in Source-to-Pay technologies that are truly AI-enhanced. (Which, FYI, might not match one-to-one with what the doctor chronicled five years ago because, like time, tech marches on.)

Today we move on to AI-Enhanced Supplier Discovery that was available yesterday (and, in fact, for at least the past 5 years if you go back and read the doctor’s original series, which will provide a lot more detail on each capability we’re discussing). (This article sort of corresponds with AI in Supplier Discovery Today that was published in March, 2019.)

YESTERDAY

Smart Search

As penned in the original, while this is not really AI in any sense of the definition, extremely powerful searching and faceted filtering can really help an organization find the information, or in this case, the suppliers they are looking for. In the early days, searches were super simple — suppliers for product X in this category. If you wanted something like “suppliers in eastern Europe which supply widgets and sprockets with a third party financial risk score of 3 or less that is ISO UVWXY certified with a maximum carbon output per unit of Y”, forget it. You’d get a starting list of all suppliers in all of Europe that supplied widgets or sprockets (and not necessarily both) and have to vet them one by one.

But, thanks to advances in processing and database tech, traditional semantic processing, and tagging, you can now do multi-faceted searches across multiple dimensions on million record plus databases in less than a second, and do regex processing of associated descriptions for key words or phrases for specific requirements not tagged or indexed. And all of the semantic indexing and tagging can be done with traditional semantic analysis and custom trained last gen neural nets (and done with very high accuracy).

Community Intelligence

Like searching, while most of this technically doesn’t require ML/AI, community intelligence that spans ratings, capability verifications, (past) inter/intra organization relationships, and buyer sentiment can be quite useful to a buyer. It’s not just a group of suppliers that seem to meet your requirements of “suppliers in eastern Europe which supply widgets and sprockets with a third party financial risk score of 3 or less that is ISO UVWXY certified with a maximum carbon output per unit of Y”, it’s a group that will actually meet your needs, and the best way to zero in on that group is to use community intelligence from other buyers who have used the supplier and can provide valuable feedback on their capabilities and performance.

Most of this doesn’t require any ML/AI at all as it just requires ratings, feedback on various dimensions, recording of products and services used, etc. Only the sentiment analysis requires the AI domain, and it’s just building on semantic context analysis, which uses semantic processing and customized neural nets to predict sentiment (to detect things like sarcasm, etc.).

That Was It, Folks!

In the early days, Supplier Discovery was overlooked when it came to ML/AI, because it was not seen to be as important as sourcing, procurement or supplier management (because you knew who the suppliers were, you just needed to manage them better). However, as the leaders realized that the best opportunity for innovation was often in the supply chain, focus switched to supplier discovery and real ML/AI worked it’s way in.

SUMMARY

Now, we realize this was very brief, but again, that’s because this is not new tech, that was available long before Gen-AI, which should be native in the majority (if not the entirety) to any true best-of-breed Supplier Discovery platform, that is easy to understand — and that was described in detail in the doctor’s 2019 article for those who wish to dive deeper. The whole point was to explain how traditional ML methods enable all of this, with ease, it just takes human intelligence (HI!) to define and code it.