Podcast: How Corbel is modernizing capital equipment sales with AI

Industrial manufacturers have invested heavily in automation on the shop floor—but sales processes often remain manual, slow, and fragmented. In this episode of Great Question: A Manufacturing Podcast, Laura Davis, editor-in-chief of New Equipment Digest, sits down with Le’ora Lichtenstein, founder and CEO of Corbel, a next-generation CPQ platform, to discuss how AI-powered configurators, unstructured data mining, and integrated financing tools are helping equipment builders modernize how they quote, sell, and close capital equipment deals.

Le’ora brings a background in structured credit and early-stage investing, and holds a BSc in Finance and the CFA charterholder designation.

Below is an excerpt from the episode:

Laura Davis: So we're going to talk a little about Corbel today, which is your company. To start, can you give me some insight on your background, how Corbel came to be, and why you've chosen this industry?

Le'ora Lichtenstein: Sure. So my background — I actually come from a background in private equity. I spent all of my career up until founding Corbel primarily on the buy side, deploying capital across early seed venture, growth equity, structured credit, real estate — I kind of ran the gamut in the private markets.

So I always like to joke that everything finance-, investment-, and credit-related is my first love, and that is very true. I kind of fell into the world of industrial equipment through my background in finance and credit. I was working as a director of private market investing at a multi-strategy hedge fund, and we were doing a lot of structured credit deals. We started looking quite seriously into a number of equipment finance companies.

And I had this one experience that really is the origin story behind Corbel. I was sitting with this equipment finance lender. We were underwriting them to help grow their origination volume, give them a warehouse line. And as I'm sitting with the team, I hear this kind of whirring sound in the background. I turn around and it's a fax machine, and there's a pen-and-paper application being faxed in. My jaw kind of dropped to the ground, and I was like, wow, how is this a thing?

Granted, this was back in 2016, so the industry has evolved a lot. But we're still facing a trillion-dollar industry where manufacturers don't really have any modern technology around offering financing solutions to their customers.

And that was really the very starting point for what Corbel is trying to do — modernize the finance experience for manufacturers and for their customers. We've since evolved tremendously into a much broader sales platform, and that was largely driven by the pull we were seeing from the manufacturers we were working with and the inefficiencies that became more and more obvious to us across their broader sales processes — not only in regard to the pure-play finance angle of it.

Laura Davis: Yeah, that's funny. The faxes — I mean, still today, so many have to have those faxes for invoices and orders. It's something I'm surprised has stayed around as long as it has compared to all the other tech that's gone away.

So there's definitely a need for this. When I was looking at Corbel — your guys’ website — and trying to understand the full gamut of what it does, there's this big emphasis on AI. And AI is in everything now. It's an everyday occurrence, but it can mean a lot of things. So what exactly is the AI doing within the system versus maybe what it can't do yet today?

Le'ora Lichtenstein: Yeah. So AI is probably the single largest leap that we have made from a technological perspective, probably in human history. It's an incredible, incredible tool, but it's really just a tool, right? It's a means to an end. 

For us at Corbel, we wake up every day thinking about, well, how can we leverage this incredible technology within the context of industrial equipment sales? What AI and LLMs have unlocked is the ability to surface data and insights from unstructured data, no matter where it lives.

So for industrial equipment manufacturers, when they're selling a piece of machinery, the technical specifications, the functionality, and the capabilities of that machine are a key part of that sales process. When a buyer comes in, and they're looking at a fiber laser machine, or they're looking at an ironworker, or they're looking at an additive manufacturing printer, at the end of the day, they want to know: can this machine do what I need it to do?

And the answer to that, more often than not, is buried in a 20- or 30- or 40-page PDF document. So what AI does in our context — we basically mine all of that data. We train AI agents on each one of our customers’ proprietary machine data, and then it surfaces that information to the buyer in a way that is super intuitive, easy to understand, and gives the answers to the questions where and how they want those answers, helping move them along in that confidence journey of understanding that yes, this is the piece of equipment that is correct for my business application.

Laura Davis: Yeah, that's a very nice thing that it can do, especially with the amount, like you mentioned, the PDFs. There's a lot of info, especially for certain equipment, that can be hard to parse through. So it's great that it can pull that to the forefront and give quick answers.

And you mentioned it'll take unstructured data. So if a manufacturer were to use this and implement it, do they need clean data and catalogs of all of their pieces of equipment? Can it take just what they've got strewn through different systems and pull it together? Or do they need to make sure all their data's in order before they go down this road?

Le'ora Lichtenstein: Yeah, that's the amazing thing about what AI and LLMs have unlocked — it doesn't need to be structured at all. Historically, if an industrial equipment manufacturer wanted to create a knowledge hub or some sort of software where their customers could come in and start asking questions, that was oftentimes a multi-month, multi-year implementation cycle that cost millions of dollars. Because the first step was taking all of that unstructured data and organizing it in a way that you could implement into the software. Now with AI, we can take data — whether it's in PDFs, which is the most structured format, videos, sales transcripts — really wherever that data lives. We just put it into a RAG database, and the AI agents are trained off of that data. So it's really accelerated the implementation of AI in this industry tremendously and created insights on top of that that are hugely valuable across the organization.

Laura Davis: I think that'll be a load off a lot of companies’ backs, that they don't have to worry about that because the LLM will just come in and do it for them. 

I'm curious — there's a huge push for custom configurations now and one-off pieces of equipment. People want it tailored to their stuff, and manufacturers are being pushed to provide that. So in this case, if somebody's using Corbel on someone's website and they're looking at a CNC and they need it to do a specific thing and be a specific size and have a specific feature that doesn't come right out of the box, can the company set Corbel up to cater to that? Or is that still something where they have to speak to an engineer instead of using the system? How is that navigated?

Le'ora Lichtenstein: Yeah. So I'll start off by saying that one of our core beliefs at Corbel is that technology will never replace human beings. It's meant to superpower humans to do what they do best.

In our context, that means letting them be experts in the machinery in a way that AI never can be. AI can get to a point where it is 80%, 90%, 95% confident in things that it has been trained on over and over again. But it needs to be data points that it has seen across multiple documents and multiple instances in order to say with confidence that that is the answer to any given question. So we're not believers in AI fully replacing the human technician or the human engineer. We're going to get that buyer to a place where we have answered their questions to where they feel confident enough to say, yes, I'm ready to speak to a human being.

The modern buyer has shifted. They want self-serve tools to better understand their options, the equipment, the functionality of the base model. And once they receive enough confidence that yes, this is what I'm looking for, then they'll route into a human-in-the-loop sales process, and that's where we'll hand it off to the expert technician to take it from there.

About the Podcast
Great Question: A Manufacturing Podcast offers news and information for the people who make, store and move things and those who manage and maintain the facilities where that work gets done. Manufacturers from chemical producers to automakers to machine shops can listen for critical insights into the technologies, economic conditions and best practices that can influence how to best run facilities to reach operational excellence.

Listen to another episode and subscribe on your favorite podcast app