Vendor Management Co-Pilot’s Role in Shaping Supplier Relations

Brian :
Well, good morning, everyone. Good evening. Good afternoon. It's great to be with everyone today. I'm Brian Kalish. I'm Principal at Kalish Consulting. I've been working with the folks at Hyperbots for about two years now. Here's the next session that we have. Again, it's a great series that we have going—getting some really great insights from really smart people. 

With me today is John Silverstein. He’ll do a little more introduction on himself. With that, John, do you want to just introduce yourself? And then we'll just jump right in.

John Silverstein :
Sure. Thanks for having me. I'm John Silverstein. I'm the Vice President of Finance at XR Extreme Reach and I’ve spent over twenty years in finance roles, including as CFO of a couple of companies. I’ve also worked in finance transformation as a consultant and now I’m doing transformation at XR. 

Brian :
That's fantastic. It's great to have you back and I thought maybe, this time around, we'd explore Hyperbots Vendor Management Co-pilot and how it's really reshaping supplier relationships. 

John :
Excellent. 

Brian :
Just to kick things off – I think one of the questions I hear a lot – I've got the good fortune of traveling around the world and talking to people and these challenges are truly universal. One of the things I hear is, why is vendor management still such a challenge for both finance and procurement? 

John :
Yeah, there are a number of pain points in the vendor management process. It’s the time it takes to even onboard vendors. Even at XR, when we're dealing with our customers and things on the other side of it, it’s near impossible sometimes to get things through fast enough for the companies and what they need to do—particularly in our space, where we’re in advertising.

The onboarding itself, because of approvals and everything, can take days just to get a PO out—sometimes even more than a month—to go through getting the W9s, bank letters, certificates. 

One of our products at XR actually exists because of how long this process takes—we speed it up for the largest advertising brands in the world. One of the key aspects we focus on is identity and document risk—making sure there are no gaps or fraud exposures.

It’s also about communication—getting things through different teams, figuring out who can make the payment versus who approves it, and dealing with large bureaucracies—especially when you’re working with the largest brands.

You also deal with issues like typos in addresses, incorrect ACH information, or ACH changes that might not even be real. Then you have to confirm all of that, which takes time. There’s also the challenge of making sure you have the right documentation scattered across inboxes for audit trails. Sometimes you have to search through all your emails just to figure out where things stand in the process.

Brian :
Great. I think what people who are familiar really want to understand is, what exactly is Hyperbot's Vendor Management Co-pilot? Because I think that resonates with people when people talk about what a co-pilot is. 

John :
Yeah, excellent question. The Hyperbots Co-pilot handles some of these tasks that are often spread across multiple people in an organization. The key is that you're compounding your existing resources with something that can speed up these processes.

For instance, onboarding suppliers—it can do that in minutes. It collects the data, the documents, verifies them, follows up, and creates a clean ERP record. So, approvals happen in a timely manner, and you don’t get rejections that force you to restart the process. It also continuously learns. Each approval, rejection, or expiry update trains the system for the next cycle, so it does even better each time.

This is especially useful for global companies, where time zone differences slow everything down. The Hyperbots Co-pilot operates 24/7. Humans don’t—we all like to think we do, but we sleep. So it’s critical to have something that continues the flow and process without delay across time zones.

It also has a self-service portal. It tracks POs, invoices, and payments. It sends proactive notifications and even communicates back and forth. That’s the key to the Co-pilot.

Brian :
Yeah, I think the two big takeaways for me are: one, the system's learning, right? We teach a person and they get better—but here, the whole system is learning. I call it an "electronic employee." Every electronic employee learns continuously. It’s not linear—it’s exponential.

And then, I hadn’t really thought about the 24/7 component—but it’s perfect. Sure, we can have people across the world, but there are still disjoints. But electronic employees really do work 24/7. Like you said, we just feel like we do.

John :
Yeah, and it’s exponentially looking at the same things all the time. It's just going to get better, like you said. Versus a person—they have good days and bad days, get tired of looking at the same thing, or miss details. The Co-pilot will follow the process every time, and that’s key.

Brian :
And, I guess as an attempt at humor—there are no HR issues!

John :
Right. 

Brian :
To get a little more technical, how does the co-pilot operate, really, end-to-end? 

John :
Yeah, the key is the configurable onboarding policy—it requires documents and sets risk rules by vendor type. So, you set up the policy. Then AI captures and verifies—extracts the W9, COI, bank proof, cross-checks IRS and sanction lists. The AI does all of that verification.

Then it sets up the approval workflow. So, low-risk vendors get auto-approved; high-risk ones get routed to compliance and the appropriate personnel. The Co-pilot also handles ERP vendor master creation—it posts the record and returns the vendor ID, so you don’t have to wait for someone to manually do it.

It really does multiple roles in the organization. It can be tracked and monitored too. It automates PO dispatch—this is key because this can take a while. It has a portal with email and acknowledgment tracking. So you can go through the PO process with real-time collaboration.

There are often delays just trying to get answers through email. The Co-pilot can chat on PO changes, invoice disputes, payment queries—in real time. You don’t have to wait for someone to research and respond. It does the research and gives you the answer.

It can even flag that a PO change is needed, or push payment processing or invoice rejection notifications, with notes explaining why something was rejected or the remittance status.

Right now, I can tell you—we’re always chasing remittance. It’s a real problem across all industries. And then there’s the hash-chained audit log. Every field change and document upload is captured. That’s critical for compliance and tracking.

Brian :
Excellent. Again, I think people are interested in understanding the hard results that can be measured. Because if we can’t measure it, we can’t manage it.

John :
Yeah. The biggest key is time savings and accuracy. One big metric is onboarding speed. Before, it could take nine days—and we’ve seen cases where it takes over 30 days to get a PO through vendor management. Nine days is actually aggressive. With the Co-pilot, you can do it in under a day—eight times faster, or even more.

Data quality also improves. Duplicate or bad vendor records drop. Accuracy improves from around 6%, error to 0.8% or even less than 1%. Processing costs for managing the vendor master can drop from 100% to 75%. So you’re getting faster, more accurate results at a lower cost. Compliance gaps should be near zero because the system checks all the rules and policies.

Vendor satisfaction improves too. Right now, vendors are often unhappy because of delays. With the Co-pilot, it’s one click—it goes through the whole process, like a project manager overseeing everything.

Brian :
Right, and it's funny because I'm a huge believer that processes can fix a lot of things. And I always highlight how important project management is. For finance people, it's one of those things we often put on the side—but it can really make a difference and have a big impact.

John  :
Right.

Brian  :
So again, we’re in an interesting time. A transformational time. People are always trying to understand what AI really is and what it can do in the office of finance. I’ve probably said this to you before: I’m a huge believer that AI isn’t going to replace people—but people who use AI are going to replace people who don’t.

So I bring that up because we’re starting to talk about how the ownership of the vendor process is shifting. To some degree, is it moving away from humans to AI? How does that relationship work?

John :
Yeah. So we can go through some of these tasks involved in the vendor process. In the beginning, to kick things off, once a vendor is set to become a vendor, there's data that needs to be collected. Typically, this is still done through email—believe it or not. It's not like there’s a centralized portal. You still have to talk to people manually, log in, get passwords—all of that.

Previously, it was mostly done via email because you didn’t need to create logins and all that. Once you move past that, after you get a co-pilot, the AI invites you into a portal and tells you where to go. Then you verify tax IDs. Traditionally, that was handled by an AP clerk. Now, AI can collect and validate those tax IDs and documents, check their expirations—previously done via spreadsheets and back-and-forth communication.

Now, AI and alerts manage all of that in the portal. You can even create the ERP vendor record, which used to be a manual entry by a procurement person or AP, depending on the organization. Now AI can auto-post that.

Then comes dispatching the PO. Currently, the buyer emails a PDF and waits for acknowledgement. Going forward, AI can send and track the PO, answer status queries—which are usually handled through phone or email. In the future, you’ll be able to just log into a portal or even speak to the AI. The goal is to make all status-related info easily accessible. AI can also send you status updates through various channels.

As for remittance advice, AI will notify automatically—you won’t have to manually check where payments are. Audit trails are crucial too. Instead of trying to track changes on a shared drive, a hash chain ledger gives you a new, tamper-proof way to track everything post-co-pilot.

Brian :
Sure. Yeah, the one that really gets me excited is when you talk about status queries. We’re moving into a world where we can just use natural language processing. You ask a question and get an answer immediately. We already have tools that can do this—won’t name them because they’re probably listening—but we all know them.

It’s fascinating. You can literally just ask a question and get an immediate answer—no calls, no searches, no manual queries. That’s a huge opportunity that really impacts people’s jobs. 

John :
Yeah, exactly. We spend so much time just trying to figure out where things stand and what’s going on—just getting a straight answer.

There will still be problems in the beginning—it won't be perfect. You can still get a bad status update from a person today. But as AI gets better at learning and refining processes, it’ll become easier. You won’t need to wait overnight or play phone tag anymore. You’ll get an instant answer and make a decision right then and there.

Brian :
Right. And with these natural language tools, they don’t say, “I’ll get back to you in half an hour.” This is something I’m personally really excited about. The technology is already here; the challenge is integrating it into the business environment.

So let me ask, from your perspective—what specific capabilities have enabled these gains?

John :
From my view, the real enablers include things like a streamlined onboarding wizard. Vendor setup used to take a lot of time—now it can be done in under ten minutes. Automated identity verification detects fraud, duplicates, and other blockers. Document validation—no need to manually check the W9, COI, or bank info. Compliance runs on autopilot.

You also have a real-time status portal. While portals already exist, there are too many of them and they’re inefficient. The ability to check your status quickly and efficiently in one place is a game changer. ERP sync gives you a single source of truth—replacing spreadsheets. PO dispatch and tracking eliminates lost POs.

Automated remittance and feedback reduce the number of status emails by 70%. So many people struggle with email overload—this tackles that directly. Flexible workflows and alerts for exceptions—these are automated now. We also discussed audit trails – with a hash chain ledger, you can pull SOX evidence in one click, instead of hunting through shared drives.

Also, unifying everything into one portal is critical. At XR, for example, we have multiple ERPs. Hyperbots supports multi-entity, multi-ERP environments with native connectors. That’s a key success factor.

Brian :
That’s really important. Excellent. 

I think a fair question we’re all trying to answer is—what is the competitive differential? How does Hyperbots stand out in vendor management?

John :
Hyperbots stands out because it’s AI-first. Before diving into features, that's the most important point. It wasn’t built as an ERP with AI added later. It was designed with AI from the start. So tasks that used to take days—like AI verification—now take minutes.

Identity and sanction checks are automated, not external or manual. Learning loop models improve with each cycle, unlike traditional portals or RPA systems that break with any anomaly.

Agentic actions like vendor ID creation, PO dispatch, and remittance are handled entirely by the Hyperbots Vendor Management Co-Pilot. Traditional portals only display static data and pricing is different too. You pay per active vendor—no seat licenses or hidden fees. The value lies in the automation, not in adding headcount.

Audit trails are secured with hash chains—no more scattered attachments. Implementation is another win—3–4 weeks versus the typical 3–6 months. And it improves continuously after rollout. Plus, multi-ERP support comes natively. Hyperbots doesn’t just let you upload files; it thinks, verifies, posts, and notifies—without bloating your license scroll.

Brian :
Exactly. That’s what people want to hear first. I love how you pointed out that Hyperbots is AI-first. That’s like the difference between a car that’s built to be a plug-in hybrid versus one that’s retrofitted.

I recently bought a car designed from the start as a plug-in hybrid. The tech works better than cars that were built as gas-powered and then converted. Hyperbots is the same—it wasn’t an afterthought. AI is the DNA.

John :
Right. Most systems today are trying to figure out how AI can fit into their existing processes. Hyperbots starts by asking—how can AI completely transform this process? It’s not just a feature. It’s a fundamentally different way of doing things.

Brian :
Absolutely. I come from an engineering background, and it’s a big difference—whether you’re building for a purpose or trying to adapt something.

Thanks again for all the insights. Do you have a final comment to sum up this discussion?

John :
Sure. 

The vendor onboarding that once took nine days or longer, sometimes thirty days to finish, can now finish in hours with a compliance risk that's virtually disappeared. The suppliers can get real-time visibility and Hyperbots Vendor Management Co-pilot makes that happen without human-babysitting. 

Brian :
I like that term – babysitting. Well, John, any final thoughts? If not, thank you very much for your time. I had a great time. I've learned a lot, thank you. 

John :
Yeah, thank you for having me and I look forward to more discussions in the future. 

Brian :
Great. Be well. Thank you. Take care, all. 

John :
Thank you.