CapTap Automates the Full File with Heron — From ISO Submission to Priced Offer in Minutes

CapTap took its first submission late April 2026. Every one since has run the same path: in by API or email, bank transactions scrubbed, cash flow analyzed, existing positions surfaced, background checked, policy applied - a complete, decision-ready file in about ten minutes. In month five, having tested Heron's read of the file against their own underwriters across fifteen hundred offers, they let a first set of offers go out with nobody reviewing them at all.

"Within our six months here at CapTap, we're able to put out automated offers in just a few minutes - which we believe is the future of this small business lending space." - Dane Steinbarger, CEO, CapTap

• About 10 minutes from ISO submission to a complete file - cash flow, positions, background, policy, all of it
• 7 in 10 applications declined before an underwriter ever opens them
• Automated offers live since month five - smaller deals priced and sent in as little as 5 minutes

Background on CapTap

CapTap is a merchant cash advance funder founded in April 2026 by CEO Dane Steinbarger and his partner, Ralph Mirabella. Both previously ran the ISO channel at one of the largest funders in the space. They fund a couple of million dollars a month, with terms of up to 12 months and advances of up to $200,000.

When Dane and Ralph began thinking about building their own MCA funder, they wanted to lean fully into AI and take a technology-first approach to the business. With their years of experience, they understood the best process for handling an application and applied AI to help build CapTap’s platform. Their deep experience in the broker channel also helped them develop the tools and partnerships they use today to set themselves apart. Dane and Ralph first saw Heron at a deBanked event in early 2025 and knew, before CapTap took a single submission, that Heron would sit at the top of the funnel.

Pain Points

CapTap designed their automation process around the one most funders still use: a submission comes in, and someone starts manually reviewing. It takes up to two hours per file just to build a baseline understanding - before anyone can determine whether the deal merits a second look.

"The prior company that we were with had an offshore team doing all of the manual data entry. I actually did that at a previous company when I very first started, and I was shocked that it's still taking place." - Dane Steinbarger

Background checks were the second manual step, and the worse one. Someone opened a 30- or 40-page report and read it to see if the person and business had any hits worth declining for.

"It took several minutes, and it left room for error." - Dane Steinbarger

The deeper the merchant's history, the longer the read and the more room for something to be missed - on a file that, statistically, was probably going nowhere.

Roughly 75% of applications reaching CapTap are deals they can't work with. Seven times out of ten it's the cash flow: not enough revenue, too many negative days, existing positions the merchant is already carrying. 30% of the times is background: MCA defaults, recent bankruptcies, large open tax liens. Every minute of hand-building goes into a file that was never going to fund, and the deals that deserved the attention are waiting behind it.

Earning Trust Before Automating with Heron

Putting Heron at the top of the funnel was the easy decision. Letting an offer go out without a person signing off was the harder one. As Dane puts it, once you send an offer, someone can take the money - and you’re on the hook to get it back. That kind of trust has to be earned.

So CapTap didn’t flip the switch overnight. They built their underwriting algorithm and kept a human in the loop through the first few thousand offers, before they let anything run fully unattended.

They stress-tested the data behind the algorithm to make sure it held up. The question wasn’t whether the statements parsed, but whether Heron caught the things that actually cost funders money: undisclosed positions, line of credit pulls, and transfers between the merchant’s own accounts that can look like revenue.

"Knowing that when the data comes over to us, that it's accurate - meaning they didn't miss an MCA position, didn't miss a deposit from another funder, they didn't mislabel some of the revenue transfers back and forth." - Dane Steinbarger

CapTap set its own metrics and knockout rules; Heron built to fit them. Captap’s self defined checks determines which submissions to decline versus review. Edge cases go into a shared Slack channel and come back answered within a couple of hours, usually with a fix attached.

"That type of partnership and trust is built over time … We've really been able to truly lean in and fully trust all of that information." - Dane Steinbarger

How Heron Powers CapTap's Underwriting

Submissions arrive by API or email and they all take the same path:

• ISO & Bank Transactions including bank statements & open banking takes the submission the moment it lands and extracts the business data - no inbox triage, no data entry.
• Industry Prediction returns a six-digit NAICS code on every file, so nobody is guessing at what a merchant actually does.
• Cashflow Analytics turns the transactions into a repayment picture: revenue, negative days, existing MCA positions and what each costs as a percentage of remit.
• Background Checks screen every surviving applicant against premium nationwide public-records sources - liens, judgments, criminal records, bankruptcies - and remove anything tripping CapTap's rules.
• Credit Policy Decisioning applies CapTap's own rules and auto-declines roughly 70% of submissions before any underwriter sees them.

"Within 10 minutes we have all of the information front to back, all of the cash flow, all of the background. They're teeing it up for us to spend a few minutes reviewing all of the pieces, all the data points, and making a credit decision." - Dane Steinbarger

What comes out the other side is a decision-ready file. Most of them still go to an underwriter; however, a growing set doesn't.

Outcomes

1. The First Set of Automated Offers Goes Out

CapTap switched on automated offers in month five, letting their underwriting algorithm decide which - starting with smaller deals. Those offers go out with nobody reviewing them - a submission becomes a priced offer in under ten minutes, against about two hours when a person is in the path.

"The idea for us is to grow and scale that number. We'll push the boundaries as we see more of these deals perform." - Dane Steinbarger

These automated offers are already funding. As those deals perform, CapTap is gaining confidence in expanding the program - taking on slightly larger advances, a little more credit risk, and new merchant profiles. Dane’s goal is to automate ~80% of their offers, but they’re only widening the box when the results support it - not simply because it’s time to move faster.

What the speed buys is position. An ISO working a deal sends it to several funders at once, and the first credible offer back is the one the rep takes to the merchant. But being early only counts if the offer holds.

"In order for speed to really be something that the partner that we work with can count on, it has to be speed, ease, trust and accuracy." - Dane Steinbarger

An offer that arrives in minutes and then gets walked back is worse than a slow one. CapTap's position is that it can be first and stand behind what it sent - which is what turns a fast answer into repeat submissions from the same ISO.

2. Underwriters Only See Deals Worth Deciding

Deals that are flagged for review reach CapTap's underwriters scrubbed, analyzed, policy-checked and background-verified, with positions and industry attached.

"All of the data is like - it's like a softball lob right up to us. So we really don't spend that much time before we're able to take a swing at that and put the offer in play." - Dane Steinbarger

The win isn't that underwriters move faster through the same queue. The queue is now more refined. Every file in it is one where underwriter judgment changes the answer - the genuinely marginal deals, not the obvious ones.

Humans haven't left the process; they've moved down it. The deeper review now happens after a contract comes back signed, where a mistake is expensive and a person catches things a rule can't.

"All of the automation on the front end means that we can take our hands off of that piece of the process." - Dane Steinbarger

3. Growth Without the Headcount

Six months in, CapTap funds a couple million dollars a month on a team built around automation instead of volume hires.

"Without Heron, we probably would not be able to be where we are today in our 6th month … nor with the headcount that we're at today." - Dane Steinbarger

Fraud was the part Dane hadn't expected to care about this early. His reasoning is portfolio-level rather than deal-level: fraud losses get priced into every deal that follows. That's why the Heron’s Fraud Consortium appealed to him. Members share the deals that defaulted, and Heron checks every new application against the companies tied to those defaults across the 50,000 submissions it sees each day, so a fraud ring that burned one funder gets flagged before it reaches the next.

"As fraud grows, we have to do more in order to combat that, because that ultimately affects every customer in terms of the pricing and the offer quality that we're able to put out." - Dane Steinbarger

The same structured data has a second use coming. Six months in, CapTap's book is too young to draw conclusions from - but every deal now carries its cash flow attributes in a form that can be queried later. When defaults start landing, the question stops being which files went bad and becomes which attributes they shared: a negative-day threshold set too loose, a position level that should have disqualified. Then the policy moves, and the next cohort of submissions run against the improved rules.

The Result: Building the Future of Underwriting

CapTap’s bet was that a funder starting in 2026 shouldn’t inherit a 2016 process. Heron enabled them to accelerate - not by speeding up an existing workflow, but by never building the manual heavy process in the first place.

What they’re building toward isn’t a headcount number. It’s a model of underwriting where the machine carries the file and people carry the judgment, with the line between the two moving every time a cohort of automated deals performs. Including pivoting back to Heron when agreements are requested to streamline the funding process - like drivers license and voided check parsing.

"That's definitely the idea for us: to continue to lean into AI and automation, to have humans just spend time where it's more valuable in our process." - Dane Steinbarger

Dane has been in this market fifteen years - long enough to watch a competitive turnaround go from hours to minutes, and to conclude it isn't finished moving. His view is that automated offers are where small business lending ends up, and that the funders who get there first will be the ones who built for it rather than retrofitted.

"We've already had a couple referrals of people that we've sent over to Heron - we're just impressed by the team, the offer, the partnership that we have." - Dane Steinbarger