Transcript: Enterprise procurement: AI frameworks for supplier data and visibility

Attention came mainstream on this topic in the pandemic when very rapidly global supply chains just collapsed. However global supply chain in the last five years has been through the stress test first with the pandemic then a ship being stuck in the swars Canal geopolitical priority issues within US and outside US with goals to bring and control more of the manufacturing for entire co is actually a company that is helping governments and enterprises can efficiently manage the supply chain administration of Farm when you know there was a huge data stream from China as a supply chain what shifts have you seen geopolitically like the geopolitics that we have in the world and rising tensions between United States um and other countries including but not limited to China you may have heard this term weaponization of the supply chain so that is where Chinese customers are now shut off from US Supply well it goes in the other direction too and the the costs of it being on self problem are being experienced in the millions and billions of dollars every single day how long did it take from like the pandemic reaching out to you can you give some concrete examples of the most like critical or surprising data points initial approach was actually to make the product initially free. First customers really came to us and that included actually at the Pentagon in the Department of Defense. Second big thing is the goal is to really have a 360 degree view of company not just its financials people tend to think of what's its revenue or how much money has it raised and our belief from early on was that that's only one piece of the puzzle really interesting data set is who works there. What's the executive team like who are they hiring and so we felt We've had customers who in the quote unquote old days would take about eight person hour well our AI now does it in about 45 seconds and we're really only at the beginning. Hello everyone welcome to startup project the podcast dedicated to encoding stories strategies and insights behind the world's most innovative startups and the people who are building them um global supply chain in the last five years has been through the stress test first with the pandemic then a ship being stuck in a swars Canal the geopolitical priority issues within US and outside US with goals to bring and control more of the manufacturing for every country multiple wars and for us every problem is a tech problem and C Co is actually a company that is helping governments and enterprises who manage their supply chain by building technology that can efficiently manage their supply chain my guest today is Leia Lev tov a brilliant mind behind the Craftco company Leia's journey has been impressive and diverse he's a graduate from Columbia University and Stanford graduate School of Business Craftco is an AI enabled platform that provides 360 degree usability and risk mitigation for supplier networks they serve Fortune 500 companies over 60 government organizations including Department of Defense Craftco has not only secured significant funding with the series B round of 32 million dollars but has also become indispensable tool for procurement and supply chain professionals worldwide in this episode we'll deep dive into building a category defining company the intricacies of global supply chain and power of building a resilience supply chain for your company Leia welcome to the show. Thank you very much it's great to be here and thank you for that fantastic introduction. Uh so I think uh you know everyone sort of uh has seen supply chain go through these ebbs and flows last uh maybe I would say five years starting from the pandemic just give our audience you know what is the lay of the land of global supply chain today from your point of view? You're absolutely right that attention really became mainstream on this topic in the pandemic when very rapidly global supply chains just collapsed and we also how brittle they were and in parallel it was a kind of peak period of of globalization uh the end of 20 years of of outsourcing to the lowest cost territories around the world all creating really thin margins really thin just in time uh supply chains that were inherently brittle and yet and the pandemic blew open uh the the the revelation that they really were were not resilient um since then however I would say that the the focus has been more secular than cyclical in the sense of it didn't when the pandemic was over the issue didn't go away um we we simply were aware that um for all kinds of reasons the the complexity in the supply chain and how incredibly dynamic all of all elements of the global economy uh um and and um you know earth's geography uh how they are they they create a dynamic in which you know supply chains are just extremely sensitive um and we're really only at the beginning I would say of truly uh mapping and understanding them thoroughly uh with with data and intelligence still very very early. We talk to I'm assuming you're talk to you know top leadership in supply chain within companies or governments what is the main concern and why they're looking into you know like adopting a tool or a software like Craftco what are the main concerns? Yeah so I'll I'll start that by saying the kind of different angles you know dimensions of of risk or or or topics with respect to their suppliers they get certain attention but then after talking through a few of those the issue really is more holistic and to do with a kind of an overall process um and approach to managing a supplier ecosystem in general so first of all what are those specific dimensions that get people's attention cyber might be top of the list uh because of uh how much we're seeing um cyber attacks on companies come in through their suppliers uh several years ago Mesk had um a a massive supply chain hack come in through one of their suppliers late tapt did the same recent quantus Jagged Land Rover these massive corporations where it wasn't their direct uh networking sort of a traditional cyber hack that that took place but it was really a weakness at one of their suppliers that then had a connection into their systems or point of sale that led to very significant costs. So the CEO it's now charged chief information security officer is now no longer just looking at is my network secure but to what extent is the network of my suppliers my key suppliers secure because if not that that can really bring in um a huge a huge problem the second uh big risk uh an area that that people are focused on and this is particularly true in government but it's also true in regulated Industries where because of the geopolitics that we have in the world and rising tensions between United States um and other countries including but not limited to China you may have heard this term weaponization of the supply chain um we're all aware of uh foreign ownership control and influence or Foi uh there's an increased focus on government-owned entities and state-owned entities because the theme here is about is this supplier of mine actually closely connected to uh an adversarial nation state and if so what risk does that bring to my supply chain the answer can be very very significant um and it goes both ways by the way we've seen um the United States uh issue sanctions and restrictions that make it impossible for a company to sell to to uh into China right we're aware of that in semiconductors uh and and other areas and so that is where Chinese customers are now shut off from US supply well it goes in the other direction too and China uh can very easily say to a supplier of let's say uh battery technology you're now no longer able to sell to any American company and suddenly you know this American manufacturer has just lost its biggest uh battery uh supplier so that whole topic of geopolitics sanctions government-owned entities and and foreign ownership uh is another very major uh risk and consideration that people have focused on uh after that I would say ESG uh environmental social and global uh and governance considerations um so how is that company governed do they pollute or not uh do they obey global Labor laws such as avoiding forced labor in their manufacturing processes these are all as well they should be still very very important topics although I would say that the focus on ESG has arguably decreased in the last year or two and then lastly there's the the most traditional of of concerns for the supply chain which is financial will that supplier be around in six months or 12 months or how financially secure are they are they so those are all those are all issues that are top of mind rightly so for supply chain leaders and procurement leaders in large Enterprises and then I I said I would um you know draw the thread together that while those are the individual some of the individual risk dimensions what's really going on is that large Enterprises have not yet really been able to evolve very advanced and sophisticated risk governance capabilities that really ingest and take advantage of all the data and all the insight and all the intelligence that you would like to have about your supply chain and in the past it was the data wasn't available uh or the pipelines were not built it's often an issue that where they are built they're still very disparate and siloed um across the organization you'll have certain sources with the finance team others with the legal with corporate development with the it team you know everybody in the Enterprise has got kind of their own tools or or data sets and it's completely not integrated uh into what you would like which is a comprehensive holistic um approach to comprehensive supplier management and supplier risk management and ultimately that is what craft delivers for large Enterprises I mean how how do you make sure I mean does this picked up like data is at different places but then there is also the integrity of the data like you know there might be a manufacturer in some small town in China or shanan how do you make sure that the data you're getting about them is particularly trustworthy and yeah it was something that every customer of yours can actually rely upon yeah it's a very good question and a and a big challenge in the space our answer to it is that more is more I mean it's in on all of the above strategy to get the most data and the best data and then validate it as well as possible one of the ways we do that is by having multiple data sources for a single type of data take um cyber security for example um we have three Premier Partners integrated into our platform in addition to our own uh native agent um data aggregation and scanning and since we've evolved to have not just one not two but three and plus plus sources plus our own uh native sources what that gives us an ability to do is compare the output and wherever you see the output is very consistent among multiple different data sources you get some higher confidence that it's accurate and yet when you see a diverent that is usually a signal or a flag okay someone's not right here we're not sure who but let's go take a deeper look um and that ability to throw a flag when something might be wrong it's not definitely wrong we don't know yet but it might be wrong and it focuses our attention is a really valuable piece of the puzzle call it triage basically a large Enterprise may have 20 000 tier one suppliers they don't know which ones to go focus and and and look at but with this methodology suddenly it's down to a few hundred that are showing some signal of maybe something being wrong now you attack you you put your human effort into investigating those keys and if it means boots on the ground go pay a visit to the supplier to get the right answer then so be it but but but the value a large part of it was in highlighting which of the suppliers uh potentially you know require um some extra expiration. You you mentioned the uh example of Jaguar and other companies which faced uh cyber attacks on their suppliers um once that happened how does craft come in and sort of prevent that in future or like prevent that and you know for your companies yeah well well we certainly can't do anything while we're not there um but but it it's quite a common story that that companies that start to work with us have faced some sort of very significant disruption um and it's brought them to the conclusion that the status quo is no longer good enough basically that world of having no data sources or disparate you know different ones but not centralized and no really advanced program of systematic risk governance a company companies are increasingly recognizing that that they need to address that and that's what you know creates the entire market that we're operating in that is already large and and growing rapidly because of these Dynamics it's a completely unsolved problem and the the costs of it being unsolved problem are being experienced in the millions and billions of dollars every single day so once we do start with a customer when we come in the first thing we do is identify all of their suppliers um and in some cases it may be a heavy focus on just their critical suppliers they might be determined by level of spend uh you know it may there's there's not going to be an equal focus all the way down to some tail indirect supplier of um you know some office supplies perhaps as much focus as there is on you know a core ingredient for the manufacturing products that that that this Enterprise creates um but we will take create a record um of every supplier and we will enrich it to make it the the most comprehensive um complete uh profile of that supplier that is possible uh through a mix of data that's aggregated by us and our agents uh that's complemented by best in class data uh from from the wide array of data partners that we have um then complemented by data that the Enterprise themselves has on these suppliers such as what's the relationship what's the spend what product lines does that supplier uh feed into and then often some data that comes actually from the suppliers themselves for example filling out surveys or uploading certificates um and so that whole array of first party second party and third party data as we call it creates the golden record of the supplier and that's the basis for the service you've now got that and anyone in the organization can access it in all of those different functions uh procurement supply chain finance corporate development legal uh it and cyber security ESG and sustainability all of those teams now have access to a shared common uh source of truth on that supplier and that's the Baseline that's the Baseline of our our data fabric and then we're elevating with AI returning that into really rapid highly efficient streamlined analytics um and intelligence for example running specific risk scans periodic risk scans on portfolios of suppliers that used to take hours and days and weeks uh can now be done in seconds and minutes uh in our intelligent workspace So uh in in the case of cyber attack example now the Jaguar or some of the company can look at their suppliers and basically evaluate you know how risky is it to work with this particular supplier because you know whether they are prone or not prone for a cyber attack is that how they're making decisions? That's absolutely right and let me expand on it a little bit more so focusing into the cyber area that we talked about so critical um suddenly with craft and the craft supplier profile there's a whole array of scores and descriptive qualitative uh attributions provided by some of the most advanced cyber scanning companies in the world as well as our own native um aggregation of of of up-to-date information all coming together to really show um the user of our product the overall risk profile of that particular supplier with respect to cyber security and so obviously the first thing that happens is you see who you're bottom quartile or desile of suppliers and you are suddenly aware that these are the key suppliers who haven't recently patched their software um or or are using outdated root software um or who have had passwords leaked onto the dark web okay just just a couple of examples of the things that you can then very very easily spot on one or more of your suppliers that can immediately direct uh mitigation efforts that's what happens on day one and then continues on an ongoing basis with 247365 monitoring such that even the supplier that looks safe yesterday and today if something happens tomorrow uh that that renders them vulnerable to a hack that's getting alerted to you now immediately so this is a very interesting space and a supply chain is not always the law people understand it work in that industry or you know you've created supplies for a particular product and you worked in like corporate finance early stage at crunchy ro you did a little bit of venture capital and how did that whole thing translate into craft that's a great question and uh yes I've got I'd say quite a diverse uh well-rounded background you might say um I started out in corporate Finance uh uh at Goldman Sachs in London I came over to business school at Stanford I had build held business development roles at two venture back startups spot runner and crunchy roll as you mentioned uh and a stint uh for several years as a venture capital investor with venrock I also held a a business development role uh focused on Enterprise Partnerships at Deutsche Telecom uh in Germany now the one thing that every single one of those roles has in common is that it had me uh every single day almost um having to do something which is look up information about a company whether I was in sales or Partnerships or Investments uh almost every day there was some company comes across the radar and I would need to look up quickly some information and get smart about that company and it seemed to me that that that experience of trying to get smart about a company um was was very uh suboptimal um you know at the time some of the sources that you had would be crunch base would focus on funding data for startups Yahoo finance would focus on financial information for public companies but that was about it for for the free sites and everything else kind of had a pay wall whether it was dun and Brad Street or pitchbook so bottom line just very hard uh for an ordinary person uh to just get good quality access to to information about companies so that was the initial problem you know entrepreneur scratching their own itch says can we build um a better platform for this digital age uh that takes advantage of all of the fragments all of the digital breadcrumbs out there let's Hoover them up and create uh the the most comprehensive advanced upto-date profile of every company in the world um we called it building the source of truth on companies um and and that's where craft started um I think you mentioned somewhere or uh I was going through some of the blog posts I think there are about um 2 100 data streams 500 data points 500 plus data points for a company uh can you give uh some concrete examples of the most like critical or surprising data points uh you know that might generate like a moments for your customers yeah yeah yeah absolutely so you're right to reference over 2 000 but 2100 individual data points they all end end point in our API and they are across 14 different dimensions of insight you know financial and operating and environmental social governance uh cyber these these are particular Dimensions and the goal is to really have a 360 degree view of company not just its financials which I would argue represent sort of the most traditional view of a company people tend to think of it oh show me the financials right what's its revenue or how much money has it raised um and our belief from early on was that that's only one piece of the puzzle you also want to know where does the company operate tell me everything you can about its products its services and its pricing a really interesting data set is who works there what's the executive team like what's the overall team like and is it growing or shrinking who are they hiring and so we we built aggregation collection of all of that data you know think about what you can extract from the jobs page of a company it's not just what they're hiring for but how that's changing over time and even what functions they do in in which locations so um Sorry go ahead Yeah so so we we've we've built aggregation for um and structuring and validation for data across uh all all all of that um uh territory and I would so I would say the most interesting uh thing for our customers has been not any one particular data point but the cross correlation across the data points that in many cases has yielded some insight that no one data stream would give you by itself for example we found a very high correlation between um high employee social media scores and engagement for example think of companies with a high Glassdoor rating there's a high correlation there between um that sort of Highly Engaged satisfied Workforce and having a lower risk of a spear fishing cyber attack and when you think about it it makes some sense that those companies where you've got an engaged employee Workforce that they're switched on they're plugged in the company's doing well people are happy to be there they're paying attention more they're not clicking on you know every link whereas these companies that have got kind of a disaffected Workforce um or things are trending negatively uh then you you often get uh a higher susceptibility to to behavioral uh type cyber hack such as um spear fishing another cross correlation we've seen you know is looking at um reduced head count um reduced job openings uh coupled with um negative sentiment and chatter on social media these three things when taken together can be a clue of of a company starting to suffer financial hardship or financial distress that can lead to bankruptcy and often we found those types of advanced signals uh giving an advanced look even at the same time as a traditional credit rating agency report said no risk nothing to see here everything's normal you know um uh no particular risk so so taking a really smart granular look and then layering these data streams together has produced some really interesting advanced insights um that that in the end of the day uh can become predictive what you said about craft reminds me about two pride which is like Bloomberg terminal which sort of also like tries to be the picture about a company and you know what's happening what could potentially happen and also in some way like crunch base does it a little bit in a smaller scale trying to find out the data points in an area where the data points are very hard to find which is about the really stage startup does no data because they are early stage right um so it's hard to paint a picture when it's too early for stage it's also interesting that you have so much intelligence data um does it mean that when I'm trying to create a new supply chain for a new product that I'm launching whether I'm a new company that's adopting craft or an existing company are you also providing a product or a service where uh you know there these are the whole options you can pick from for your new product or a service is that also something that you guys do to some extent yes you're talking about supplier Discovery and certainly by uh you know the volume of companies we've got which are millions uh and they're all tagged by uh category and a variety of industry codes um so it's very possible to look in there and see for anyone given product line yes alternative suppliers that you may want to utilize or start a relationship with especially in the event that your existing supplier is showing some uh you know risks or concerns that make you think aha I may not want to be uh sole sourced and dependent on that supplier alone what uh shifts have you seen geopolitically like the obvious shift has been I think I can see like the two significant shifts that happened one is um I think during the first administration of Trump when um you know there was a huge derisking from China as a supply chain so that uh um you know companies were moving away from China that meant US companies move to Mexico and there's a lot of diversity in terms of like which Southeast Asian country benefited from it like do you have a specific insight on like which countries actually benefited from this kind of shift and also how much of that actually moved to the US? Yeah well Vietnam is one uh you know in particular uh because a lot of changes were driven by the tariffs when that all um you know really hit the fan um in April one of the first things that happened is companies all trying to figure out work around and how can they you know quickly uh and very um uh covertly get the supply into countries that didn't have as high uh tariff rates and so there was a lot of focus on on China moving to Vietnam um uh you know in the early days of the tariffs I mean the whole thing has been so volatile it's been almost impossible to fully keep up with um it's almost like timing the market which you you know the stock market which most people agree is very hard uh not even worth trying um you know likewise um many Enterprises that we talk to take a more global and holistic approach and they say look I've got to optimize the supply chain over the next 10 years not the next 10 days or weeks and so what does that mean uh the tariffs will come they'll go they'll go up they'll go down you know that's the whole idea uh as um I think of them uh to to be volatile and to to create Dynamics where um where things stabilize in a position that's more favorable to the United States that is that's at least um you know this administration's objective with the tariffs as as far as we can tell um taking a longer term perspective for sure near shoring and friend shoring are important Dynamics which is you know the also known as decoupling uh or derisking in particular from China but elsewhere around the world I mean there's also here a build American right um sort of uh philosophy and and and Mantra which we think is is tremendously uh exciting so um you know bottom line it's very Dynamic it's an environment in which every Enterprise really must and very much is thinking almost from first principles again about where should my supply chain be where do I want it how can I have as much of it as I can uh more under my control rather than less closer rather than further away and and as we see it that that's probably a good healthy uh dynamic um for for most companies to think about why do you think Vietnam benefited like is it uh language is it Workforce like what were the reasons do you think Vietnam really you know got benefit most of all these countries? You know that that's a good question I I'm no expert there but I would say that it was simply uh a good business environment I mean Vietnam has has had one for many years they've done a fantastic job of um you know having high quality skilled labor and uh the ability to operate with trust so I don't think it was hard um for Vietnam to be a beneficiary and uh American and European countries uh to increase just turn up the volume on what what they're doing with with Vietnam while of course the point was it's extremely close to China and was able to you know just just very quickly get that supply across the border um um you know so then it started to look as if it was being exported from Vietnam and not from China now then it's a game of cat and mouse right because the authorities start looking into it and saying well where did this really come from and that's something that Customs and Border Patrol here cares a lot about and rightly so um so yeah you know it it's cat and mouse and again I would think that most of the smart focus is on the longer term uh Dynamics and and how to land the the new supply chain um in in a secure and resilient position as possible unlike you know the last 20 30 years brittle and uh and just in time um and so though that's just a very major strategic initiative over many years again not a matter of days or weeks or even months I mean uh I think with such rich data and insights or you know different diverse set of data that you have in your platform uh I mean it should be a great opportunity to leverage AI both in terms of like how you can enrich the data um and the data sources and you can have more data sources and enrich more details on each of the dimension that you're looking at and at the same time do a lot more with that data itself uh so can you talk a little bit about Yeah you know what's going with AI now? Yeah Yeah you're absolutely right I mean our business uh started um you know building a source of Truth on of on companies and collecting and structuring all of this data long before the uh the the recent explosion um in AI from from gen Ai and large language models um it's just turned into a dream scenario because what we've got in house now is a tremendous proprietary data set um made up of many many painstakingly fused together sources with a lot of entity resolution resolved a very high quality operation for data validation and keeping the data up to date um and now we've got these incredible capabilities uh coming from llms in order to uh generate inference and insight and intelligence um from all of that data so it's very very exciting you know the last couple of years and this this year in particular we've really accelerated our development um with an AI front end that a craft we call the intelligent workspace and it's it's focused on the the the procurement professional the category manager who's responsible for the relationships with a group of suppliers um and and first of all the uh the the capability that that category now manager now has in the craft intelligent workspace is is a tool working for them 247 but when then they come into work in the morning they've fundamentally got at the top of a a list top of an inbox uh here's here's three things that happened overnight with your 100 or couple of hundred suppliers that you may want to take a look at here was the signal of a company in financial distress here's the signal of us of one of your suppliers vulnerable to a cyber hack here's one that's just shown up in a court case related to a um a a labor violation you know these are sort of needles in the Haystack that category manager simply had no ability to to be on top of previously and now they're being brought right to the surface um together with all of the data and the underlying evidence um for the category manager to make a decision is this material or not so that's one of the things um all completely AI driven by just continuous monitoring and sort of intelligent scanning for all of the data um the the second big thing is these uh the moments which um our users are responsible to produce uh a comprehensive uh report on the state of supply chain um security or risk I mean that is increasingly what's happening where the chief procurement officer Chief supply chain officer CFO and then even CEO and board are really saying what is the health and the resilience of our supply chain um how much risk is there in there um and and are we comfortable with that and and how are we adjusting it so when it comes to quantifying that and creating reports against it these are the other things that AI accelerates tremendously and just to give one data point we've had customers who in the quote unquote old days would take about eight person hours like a full working day uh to create a a deep uh risk report on a given supplier covering all of these uh parameters well our AI now does it in about 45 seconds and are you are you primarily developing models within your company uh or do you see down the line developing those or are you primarily using a particular vendor or are you fine-tuning what are the types of things that you're doing yeah we're we're utilizing um an array of the existing llms at the at the moment uh you know our own models um it's not something we're focused on at the moment but I can see over time us evolving our particular own supplier risk models um especially when it comes to our desire to configure those models according to the wants of a and needs of a particular customer right so different customers in different segments care more or less about different risks and they have different risk tolerances and thresholds right that's a that's a great example of what standard llms um you know will will not do off the shelf um and yet our application very much can um that's the direction of travel for us our own models but um today it's been it's been working with uh open AI and anthropic and and Google Gemini um those three in particular to leverage all of their um um all of their AI models and run those on our particular data set and with all of the context for the use case um that that we have uh inside craft uh together with our customers what was your initial sort of go to market approach of like yeah how how did you convince the first five customers because yeah this is like a huge problem because to solve it at scale it could take a lot of time maybe a couple of years but you still need to run a company and try to sort of like validate your use case and make sure that you're not building something that will no one will pay for uh so what was that initial approach of um you know finding those first five 10 customers? So our initial approach was actually to make the product initially free um not the first company to have done that right if you think about product LED growth so in those early days when we were doing our initial aggregation of data on the first several hundred thousand suppliers we put the data up for free on long deep profiles of a company similar to a crunch base profile or even a Zillow or Trulia profile of a house very well structured um data nicely visualized and laid out and then what happened was that Google um started uh crawling the site aggressively giving us very high page rank because they could see that we have high quality well validated data so there was high trust in the data and our website actually grew completely organically to appearing over 100 million search results per month and bringing uh almost 2 and a half million people onto the website every single month um to to consume some information in relation to a query um about a particular company so it was really from that footprint out on the broad web that um that we that that our first customers really came to us and that included these first large Enterprises coming to us um with a need in this supply chain and this procurement function um so it was in Aerospace and defense uh initially that we were approached first by Prime manufacturers and then by teams um actually at the Pentagon in the Department of Defense or the Department of War as it's now called um finding our data uh on these free profiles after they did a Google search and then and then being impressed with what they saw and getting in touch with us that's a quite that's quite an inbound story how long does did it take from like the Pentagon reaching out to you uh to sort of like closing the card ride and they started using the product because it's often like you know working with government is hard and um you know it takes a long time what was your experience of uh getting that done yeah we know it you know traditionally to a government you know sales typically take very long uh and for us uh it was just a particular Dynamic this was early 2020 that that our first connection happened with the government uh the pandemic was just in the early days of kicking off but the need that they had was so acute and in their opinion the fit with what we happen to have built was so strong that we were able to close our first contract um and this was for more than a million dollars in annual contract value um in in 94 days still over three months from initial inbound conversation to sign purchase that you were over delighted at that thrilled um it was proof that government doesn't always move slowly doesn't have to when the need is there and the fit is there they can move very very fast and they did and what's been even more exciting is that since then we've grown that relationship um to over as you said 60 organizations across 26 major agencies mostly defense some civilian um in the United States federal government and um and it's uh it's a whole uh platform and service that we offer there that that we're incredibly proud of um and and proud to deliver to to help serve their mission can you can you talk about uh you know your business development and the was the size of the business today um you know I think it's almost a decade of starting the company now or um and you know give us a numbers some numbers that you can share and talk more Yeah absolutely well we're about 70 people today um with a with a a little over two thirds of that in um engineering product and design uh so so the development function um and the balancing little under a third in in our general go to market um so sales and implementation and you know little bit of shared Services operations such as Finance and People operations so we see ourselves as very lean um and uh employing more and more AI agents uh to do this work by the day did it did like AI accelerate your product development are you know go to market or general business development without hiring more people or seeing improvements within the company Absolutely and across the board uh across every function um so it started with with the coding as all of our Engineers started adopting um the the you know llm coding tools um not necessarily to write the code but certainly to to check blocks uh do validation and and it's just constantly growing there but also accelerating the go to market function researching Prospects um qualifying uh building out sales materials uh building out marketing campaigns absolutely every um aspect of go to market also um we're just using AI every single day you know back within our you know core product AI is having tremendous impact it's enabling us to collect structure and validate more data more quickly and more cheaply uh than ever before and and that trend will continue and then as I mentioned in the AI product description intelligent workspace it's allowing us to get to answers for our customers dramatically faster and uh again produce insights in seconds that previously took uh days or weeks what does the product roadmap look like um you know maybe down two three years down the line how does craft going to look like uh and what is their helpful in general for such even supply chain as an industry? Yeah yeah we're we're we're very excited about about that broadly the the theme with us really is supplier intelligence so it's being a layer of intelligence that runs across the entire procurement supply chain function um enriching awareness visibility and monitoring on every single supplier that the Enterprise has um now the interesting thing about that is that it can expand even further at the end of the day why should the intelligence layer just stop at supply chain we're really passionate about collaboration and driving collaboration and reducing silos in the Enterprise and even in this function around procurement and risk governance um where it touches the CEO and the sustainability org and finance and risk and line of business um in addition to supply chain and procurement the ability for our application to reduce silos and enable all of those people to log in um and see the same shared truth um relating to a supplier or a particular commodity or or some ingredient or component that the Enterprise relies on um that's that that's really important so the collaboration and communication layer that that we've already started building we expect to see a lot more of that what we also have a lot of on the on the radar on the road map is um data egress and Ingress so taking not only our data but our intelligent and enabling to transmit it across to another system that the user is working in um whether that's the orchestration layer or a CRM or a spend management system uh or an erp um we we also love bringing the intelligence right there to where the user is working and not require them to log into another portal um and then likewise or reciprocally receiving data from those systems is an important part of making the insight and the intelligence even more powerful I think um that's a good note uh to end the conversation it's been a fascinating uh episode for me personally and to understand what's happening in supply chain but one last question before we end the conversation what is being your favorite uh sort of way to use AI a specific tool workflow uh uh that that's your favorite use case sort of you know that has been impactful in your own life for me I mean I'm I'm just obsessed at using the research capability to get smart ahead of a uh customer uh prospect conversation right um as the founder CEO I'm still heavily involved in in sales and business development um which I love to do and so ahead of every meeting