The Future of Supplier Management (Shangyanyan Li)

Full transcript

Welcome back to the podcast guys. Today we are joined by Yan, co-founder at Bruce AI. Yan, welcome to the podcast. >> Thank you. Happy to be here. >> To start, could you share something about your background, what you and the team are building at Bruce AI? >> Yeah, totally. So, my name is Yan Yan. Originally from China and right now I'm based in Madison, Wisconsin, uh in the States. So I mean I have some experiences like working on early stage start out either as a product leader or work as a co-founder. So I have experiences like in multiple verticals including like insure tech, fintech, property tech and right now I'm building a product a venture called Bruce AI. We started working on Bruce AI in last year January uh so January 2025 and then the venture is incubated by a venture studio uh called 1848 ventures.

We officially launched a product last year in early October I want to say and then as right now we are running a closer beta program uh with a few uh beta testers. So uh we are a AI power solution uh for food and beverage industry. So we are primarily helping them to simplify and streamline supplier compliance and that's basically what we are doing with our solution. Food and beverage companies like manufacturers, distributor and importers when they are buying and working with suppliers they need to make sure they're compliant right so uh they need to collect documents when they're on boarding them they need to manually verify them and then they need to make sure all those documents are up are up to date. Uh that's how we are using AI to help them to streamline the process and save time. >> Nice. And what was the moment you actually realized that compliance and supplier management was a big enough pain in food manufacturing to actually build a whole company around it? I mean first of all food and beverage industry is a huge uh in the states I mean probably globally and also we assume like everything on the table is safe. It's actually being worked on by a lot of like food safety and QA professionals.

But like once we started talk with them like we quickly realized first of all they don't have a lot of tools uh they can use particularly for small and medium size like food businesses a lot of times like they are using Google drives right and all the spreadsheet to manage all the compliance documentation and then they use email to contact their suppliers to collect all the documents like a average customer of ours and they work with like around 50 to 100 suppliers that's a lot of suppliers to manually manage right And then for each of the suppliers when they're on boarding them they are collecting I want to say like from 10 to 20 different documents like that's hundreds of pages of documents they need to manually review as well. So like for example all their report right they need to make sure there are no major findings within their suppliers right they also need to look at all the registration certificates like from their suppliers things like that to make sure they are meeting the standards so it's just like a lot of manual work I mean when we are talking with uh food manufacturers we heard a lot of like horror stories right sometimes like they fail audit when they got visited by FD auditor uh they find out they don't have a document they need and then they got a warning right sometimes like uh when they're ready to manufacture and then they find out they're missing some kind of documents from their suppliers.

They need to pause production just like to wait for the paperwork to be in place like all that kind of stuff and we even heard like a full safety um leader got fired because of because they failed the due to lack of uh paperwork. So that's why it's a it's a big problem and people are just like doing it manually today and we think we can use AI to help them to solve it by better. >> You mentioned that there are a lot of food manufacturers Do you reach out to some specific kind of food manufacturers or how do you filter it? >> First of all, we are targeting small to medium side um food businesses. So that would be like around I want to say like 30 to 100 employees and a lot of times um they are working with around 50 to 100 suppliers and also there are like certain types of uh food businesses are definitely like see supplier compliance as a priority. So for example like distributors right and importers.

So when they are importing something into the US I mean they need to follow the foreign supplier verification program like meaning they need to get additional paperwork right they need to evaluate the risk of the suppliers they're buying from and then like there are certain types of food businesses handling higher risk ingredients like for example produce and things like that. So those are the verticals we are focusing on at this point. And when it comes to implementing the AI, could you briefly explain what kind of systems do you install in order to actually make it running smoothly? >> Yeah, totally. I'm not a CTO, so I'm not a technical person. Uh I'm going to talk about it uh from a operation perspective. So like that's the vision and end goal for us would be build a fully agendic solution uh for the food and beverage industry to handle all the compliance tasks like for example when they are collecting documents today they need to email their suppliers right and then they receive the documents and they look at them and then if there are anything missing they reach back out to their suppliers to get all the proper documents and then sometimes like for our customers I mean they may get requested by their own customers like retailers or distributors wholesalers they need like those kind documents as well, right?

So they need to like go to their Google drive or folders trying to find all the supply documentation and forward go to their own customer. So we want to like automate that all right. So in the future like we think our users don't even need to touch it anymore. So everything will be collected, verified and validated automatically and when whatever is being needed right either by their customers or by the auditors like whoever is needing those documentation I mean it could be shared automatically but as today it's not fully agendic uh it's a AI powered SAS we try model with full safety documentation so our model understands like all the languages all the requirements within the framework and also like we build a workflow we co-develop the workflow with our design partners to make sure like this is something food safety people are familiar with and can use seamlessly and then lastly we are using AI for a few things at this point.

So for example we use AI to automatically categorize like all the documents being uploaded and then we can also validate certain types of information within the document like for example is the document expired or for example is there of valid a signature in the document but those are things we can use AI to automatically validate and also we extract key information from those documentation so the user don't need to read like 80 page audit reports anymore so we just like surface uh the information they need to properly assess the risk of the supplier and And last thing I want to mention another thing is we use AI to automatically fill in a questionnaire. So for example today a lot of our customers come a questionnaire for all their suppliers. Sometimes it's like 20 questions like for some of our customers like a 100 questions. The suppliers really hate those kind of like long questionnaire.

So uh we use AI to automatically uh extract information from the documentation upload and then we can complete like 95% of the questionnaire for the supplier. That's what we use AI for at this point. But eventually, like I mentioned earlier, our vision would be make this fully agentic. So the full safely people don't need to step in uh if there's no reflex. >> Mhm. And is it a vision for the next year or is it is it something that you are looking to build in few years? >> Yeah. Uh it's probably a few years. first I mean like the technical constraints uh there are also regulations right and right now all the documents needs to be reviewed by a human so that's something it just like required today I mean even though the AI might be able to like technically validate the document all by yourself but there still needs a signature or like a log from a fully person to say like this document looks good I don't think that's going to change like anytime soon but a lot of agents like eventually want to get there >> what would you say was really big challenge for you at Bruce AI.

So there are multiple challenges right I mean we are trying to get to the market as soon as possible so we got start to learn one thing we learned is particularly for small and mediumsiz businesses and they don't want to talk with a salesperson or talk with me right so they just uh we make it a fully um self-s serve to start with even though we make the workflow super easy people encounter challenges right so they they may have uploaded the wrong documents or just like feel the workflow is too confusing so for the customers like uh we actually uh step in um to help them to set out the system properly. So in that way I mean they can have a much better experience so they don't abandon the product prematurely because like they don't fully understand like how to set up the system in the right way all by themselves. So that's like one thing we learned very quickly is like hey just give it to the customers it's not a bad idea even though we think it's simple they need some help right and then the second thing I would say is uh like suppliers they get notified um through like a lot of different portals right Bruce CI is now the first solution out there in the market why supplier told us right they have like 40 portals they need to manage some of them are tools like Brucei and then some of the retailers or wholesalers they have their own portals so they keep like receiving getting all those notifications like emails uh from all those platforms.

So they got burned out, right? So a lot of times they just don't want to open another link and upload documents to a different platform. So that's another thing we learned uh during the initial pilot. We need to do some handholding with the suppliers to make sure they upload the documents to deliver good result for our end user. So that's one thing we learned very quickly is like hey we shouldn't assume like if we build a good product suppliers will just like corporate we need to give them motivations incentives and things like that. So uh those are like a few things we learn about the product and then from a go to market perspective like right now we rely very heavily on content marketing and also like referral partnership and affiliate marketing and just simply because nobody wants to be early adopter for compliance of software that makes sense. So we don't need to build a trust uh within the industry. We feel like working with referral partners is the best way to do it. >> Yeah.

I wanted to touch up more on how do you think your solution is different than other? What's the edge that you have? >> We are targeting first of all small to mediumsiz food businesses. So we do generally feel like they are assessed if you are familiar with the regulations uh within the states. So like for all the food businesses when they reach like certain size I the regulation the same right no matter how big they are no matter how many suppliers they manage they need to follow all the rules and regulations but a lot of companies we are working with I mean they cannot afford enterprise a lot of them are like the starting tier is like $20,000 a year and then they lock you to like a multi-year contract so that's like kind of kind of pretty standard so our customers they are like hey we need something like this but we need we need it to be a little bit more lightweight and a little bit more flexible.

It's very easy to start uh with Brucei. If we get all their documents, uh we can get them started within a few hours uh once they sign up. Uh if they work with incumbent solution in the market, I it's very normal to take them like four or six weeks to get the product up and running. So the onboarding process is very cumbersome and long. And then we build Bruce to be very user friendly and very easy to get started with. And then we are AI native, right? A lot of the incumbents in the market, they are trying to introduce some AI features, but it's not as user friendly as what we offer. >> And when it comes to your role at Bruce AI, what do you currently do in the day-to-day? >> I would say I split my time between product and also go to market.

So I spend like half of my day going to trade shows like try to network with people, doing sales calls, right? and trying to work with my fractional CMO uh to make sure we generate uh leads and things like that. And then the other half of my time I work with my lead designer and also like my CTO to make sure uh we scope the product in the right way and we deliver the feature and also at the same time like working with our early customers beta users trying to get feedback from them to understand like like what we should prioritize and what we should fix. Last one. For people listening who want to check out either you or Bruce AAI, where should they go? >> They can go to bruce-ai.com. That's like our official website. Like right now we closed our uh private beta because we reached the amount of customers uh we were uh looking to recruit. We do have a waiting list.

If uh you are a food and beverage business who are struggling with managing suppliers and supplier compliance, yeah, sign up for the waiting list and as soon as we open up uh public beta, we will re reach out. So, >> perfect. Yan, thanks again. I will add the links so people can check you out and thank you guys for watching.