Joshua Burgin is the Co-Founder and CEO of Blitsy, an AI platform that helps private equity, debt and accounting firms close more deals by cleaning their accounting data. Blitsy customers include Statista, Nitinol Capital and Meijer. Before Blitsy, Joshua gained experience across corporate finance, private equity, private debt and startups, ultimately leaving the traditional Chartered Accountant path to build a company firsthand. We sat down with Joshua to talk about where AI is actually changing the M&A process, what financial professionals are still wasting time on and how technology could reshape the way deals are diligenced.
Westgate Partners: You were on the path to becoming a Chartered Accountant before deciding to leave and join a startup. What made you realize you were learning the wrong skills for what you ultimately wanted to do?
Joshua Burgin: It was never that I was learning the wrong skills. When you learn bookkeeping and accounting, you learn how to curate and analyze financial data, and accounting is fundamentally a data function for a business. It is what allows you to make better decisions about how to run your company. Those skills would have made me a better operator either way. What changed was my view on sequencing. I knew the best way to learn is by doing, and specifically by running a company, so I wanted to compress that timeline as much as possible. If I started now instead of waiting another three years, I could compound those skills far faster.
WGP: You describe Blitsy as taking QoE reports from six weeks to three. What does that time compression actually change about how a deal gets done, and where does the extra time cost sponsors more than they realize?
JB: It changes different things depending on how the deal came to them. In banked deals, speed is competitive. Compressing the timeline increases the chance that our clients actually win the deal over another firm at the table. In proprietary deals, there is no race, so the benefit shows up on the other side. Our clients get to start thinking about the value creation stage much earlier than they otherwise would.
WGP: Before building Blitsy, what made you realize that financial data preparation was a problem worth solving rather than just another annoying part of the M&A process?
JB: When I was working in the private debt space, a partner walked me through the whiteboard math on a leveraged buyout of a telco company, and my mind was blown. One plus one equals three. Then it came down to the financial due diligence behind that same deal, and I was manually copying and pasting a thousand accounting line items because the financials were in PDF form. That did not make sense to me. Why does someone with decades of finance experience have to do janitorial spreadsheet work?
WGP: You've said M&A analysts can spend hours simply cleaning and mapping financial data before they can actually analyze a deal. Why has this remained such a manual process for so long?
JB: Because the way we interact with computers has stayed archaic for a very long time. Imagine you could simply think a question about financial data and know the answer. We have never been able to translate thought directly into arithmetic and process data that way, so we resorted to rudimentary logic, if statements, to interact with computers and change bits. That is the basis of how Excel runs: deterministic formulas that can only be called because they are set functions. That is excellent for predictability, because the output never varies. It is not good for a human trying to manipulate large volumes of data quickly.
WGP: You're essentially asking AI to take messy financial information and turn it into something an investment professional can underwrite. Where does AI perform surprisingly well today, and where do you still need a human?
JB: AI is exceptional at taking a large, complex task and breaking it into smaller tasks to execute against. We have already seen it replace narrow roles in the market, the data clerk being the obvious one. Where it still falls short is nuance. When the picture is incomplete and there is not enough data to analyse, you need to press further. EBITDA adjustments are exactly that kind of work. The other gap is accuracy. Out of the box, AI still hallucinates. That is acceptable in the everyday world with a chatbot. It is not acceptable in finance, where a material misstatement has knock on effects, because you are paying a multiple on adjusted EBITDA. Data that is 100% accurate is still something AI does not deliver on its own.
WGP: You've watched sponsors run diligence from both the accounting seat and now the platform seat. What is the most common thing they get wrong about how to actually get value out of a QoE engagement?
JB: Some independent sponsors will not commission a full, deep QoE. They scope something narrower, usually to cut costs. My view is the opposite. You want to pay up front for more depth in your understanding of the company, including a site visit to genuinely look under the hood. You would rather spend an extra 10, 20 or 30 thousand dollars than make a million dollar mistake.
WGP: Big Four accounting firms have dominated QoE for decades and are now scrambling to bolt AI onto legacy workflows. What is the fundamental thing they will get wrong that an AI-native platform like Blitsy will get right?
JB: Enterprises have been poor at implementing AI into their business models, and the reason is structural. AI is non-deterministic. It works on a probability of outcomes. Enterprises are built to be deterministic, with clear cut processes designed to make sure there is no variance in the output. AI flips that on its head. Building a product that puts real guardrails around AI and leverages it in a targeted way is a genuinely difficult task. The way the large accounting firms are set up does not accommodate it, because it introduces too much variance into their output.
WGP: Most LMM sponsors and independent sponsors do smaller deals where the QoE cost is a bigger percentage of the deal size than it is at enterprise scale. How does that change what a smaller sponsor should actually be asking for from a QoE engagement?
JB: The QoE really depends on the requirements of the investor or the lender, and on the mix of debt and equity in the deal. What gets scoped follows from what those parties actually require, so the starting question is always who is relying on this report and what they need to see.
WGP: How often does a QoE finding actually change a deal outcome, whether repricing, restructuring or walking away, in your experience?
JB: Most of the time the QoE does change the nature of the deal, usually on price. Sellers are naturally interested in inflating the value of their company, which means presenting a higher adjusted EBITDA. Our job is to do the site visits and press management on every adjustment they have made, so that what comes out the other side is a fair valuation that also serves the buyer's interest.
WGP: You've talked about PE firms manually reconciling data across different ERPs and spending days building ARR bridges and board metrics. Why is financial-data standardization such a difficult problem inside portfolio companies?
JB: Every portfolio company arrives differently. Each one might run a different ERP, use different management accounting and reporting methods and have its own finance team. Where proper accounting procedures and controls were not in place before the acquisition, the reporting and the chart of accounts will not line up with anything else in the group. You need a way to compare apples to apples. That is what lets you make better decisions for the companies inside the group, understand enterprise value across the group on a consistent basis and keep the firm exit ready.
Joshua can be reached via LinkedIn.
