00:02I'm Allison LaForgia with HousingWire, and today I'm sitting down with Phil McCall,
00:08President and Chief Operating Officer at ACES Quality Management. Let's jump in and talk about
00:14the launch of ACES Population Testing. Congratulations. I'm so curious. Can you
00:21explain in plain terms what ACES Population Testing is? Absolutely, Allison, and always
00:28great to be talking with you. Population Testing systemically evaluates a specific portion of a
00:36client's population of records. That could be up to 100% if that's what they choose. This is against
00:41configurable, compliance, operational, and data quality policies, and it identifies risk indicators
00:48that warrant a closer look from their own team. Every record, every policy in real time, that's
00:56the whole idea in one line. Now, Phil, how is this different from the sampled QC audits that your
01:02clients already run? Well, traditional quality control, and it narrows the portfolio, the
01:09population, to a defensible sample, say 10% or the likes, before it reviews anything. You pick that
01:16slice, you go deep, and you defend the sample rate. This process is critical. It'll always be a part of
01:22the Defensible Quality Program. However, it only reports what you found inside that sample. The
01:29question is, what happens to everything outside the sample? Population Testing, it evaluates every
01:35record in the population. A client defines it against any design policy. Every time, the source data
01:41refreshes and reports where to look next. It measures first, then it narrows, which is kind of the opposite
01:48order of a traditional audit. Let's be real for a second. When some servicers hear every loan,
01:55every rule, they immediately worry about getting buried in exceptions. How should they actually think
02:02about that? That's a fair worry, and we've seen it happen. Our advice really comes into, this is not
02:08intended to be an all-or-nothing type of review. You don't have to test 100% of your population
02:13on day
02:14one. You define the scope, and it could be 10%, 20%, 30% of your portfolio. It could be just
02:19a single
02:19business line or a specific set of policies, and your team can review and remediate at a pace that
02:25actually works for them. The point isn't to overwhelm anyone with findings. It's to give you
02:31a dial instead of a light switch as your team gets comfortable and builds out the remediation
02:37workflows, and then you just start tweaking and turning that dial up. And when we see many of our
02:41our clients, we're all the way up to 100%, because that's how they operate. Most organizations,
02:46they're going to want to start small and then prove it out first. And both are the right answer. It
02:51just depends on what you're ready to act on. Why does the timing matter, and what does the data tell
02:57you? I think the first I'd look to is, look at our numbers from our defect industry reports. In our
03:04Q4
03:042025 trends report, the industry-wide critical effect rate moved from 1.31% in the first quarter up to
03:131.79% by the third quarter, then back down to 1.38% by the end year. So a
03:20sample pulled in Q1 would
03:21have told a very different story than one pulled in Q3. So after three straight years that we've had of
03:28improvement. We were at like 2.15% in 2021, down to 1.51% in 2024. The full year
03:37rate actually ticked
03:38back up in 2025 to 1.55%. The risk isn't concentrated in one place either. We got legal, regulatory, and
03:46compliance defects. They made up roughly a quarter of all the critical findings in our latest report,
03:51with income, employment, and asset-related defects close behind. Well, that's a lot of ground for any
03:56sampling program to try to cover consistency, consistently. And it's exactly the environment
04:02population testing is built for. Now, does this replace the human-driven audit then?
04:10Great question. No. And that's an important distinction. So population testing produces risk
04:17indicators. It doesn't produce findings, and it doesn't decide what those indicators mean.
04:22This is proactive system-driven testing that allows organizations to manage their
04:26portfolio in an exception-driven basis. You know, our clients' teams will investigate,
04:32prioritize, and determine the outcome. That depth and breadth together give you a more complete view
04:38than either one alone. Now, population testing, it adds visibility into the part of the portfolio
04:44that an audit never had time to touch. Now, Phil, AI is showing up everywhere in this space right now,
04:52including tools that claim to replace QC teams outright. How is ACE's approach different?
04:59I'll draw a hard line between AI that replaces judgment and AI that removes the work standing
05:06between your team and their judgment. We're squarely in the second camp. Population testing evaluates
05:13records against policies your teams define. And every result is explainable. You can trace exactly why a
05:20record was flagged back to a specific rule and a specific data. Well, that's what makes it defensible
05:26in front of an examiner or an investor. I'll tell you, a black box that says, trust the model.
05:32It doesn't hold up in a regulated environment, and it shouldn't. The other piece that people
05:37underestimate is depth. Anyone can stand up a rules engine. Now, what's harder to replicate is 30 years
05:44of curated compliance logic. I mean, we're talking over 25,000 checks pulled from 80-plus regulatory
05:49sources. That's not something you build in the product sprint. It's content asset as much as it
05:55is software. And that's what makes the results something you can actually stand behind.
05:59Now, Phil, where does the technology come from, and how does the workflow actually run?
06:06Well, it's technology. It's been built on top of what we did is we acquired BaseCap Analytics in 2026,
06:14just over this year. And we paired that with our ACES platform, where it gives customers that complete
06:21confidence that they're focusing on the issues that matter most. The workflow is simple. First,
06:26connect your portfolio, your production data. From there, the system evaluates it against policies that
06:32your team defines, and it identifies the records and the patterns that warrant attention. Then,
06:38using your own people's expertise, you can investigate, prioritize, and respond to those
06:42results. The evaluation itself, you know, it's deterministic and policy-driven. Every result is
06:48explainable, and that's what makes it defensible in a regulated environment. What kinds of questions
06:53is this meant to help answer? Start with the questions sampling can't answer consistently.
06:59On the compliance side, where to define requirements point to potential risk? And are indicators isolated,
07:06or are they systemic? On the data side, which critical fields are missing or inconsistent? And can
07:12your teams actually trust that data? On servicing and origination, it's things like, are required
07:19activities happening when they're supposed to? And where are the risk indicators clustering across
07:24production? Those are portfolio-wide questions. And until now, most teams could only answer them for
07:30whatever slice that they had time to sample. As an example, change detection. If a critical field changes
07:39mid-loan life cycle and, say, a flood zone designation gets updated, well, that should immediately raise the
07:45priority on that record. We've seen real cases where an inaccurate flood zone change led to flood insurance
07:52getting dropped when it shouldn't have been. And that's exactly the kind of thing a monthly sample
07:57can misentire of. Because the record that mattered wasn't in the sample the month it changed. Population
08:03testing, that catches it because it's evaluating everything, every time the data refreshes. Not once
08:09a month or once a quarter or just a random sampling that you had time or what you could manage
08:13that cycle.
08:15This interview is dropping right as everyone is attending the MBA Risk and Compliance Conference.
08:20What do you want people to walk away with? And will we see ACEs there?
08:24Yes, absolutely. ACEs will be there. And population testing, it's going to be front and center in what
08:29we're showing. What I'd want someone to walk away with is pretty simple. Every QC program, it has a blind
08:36spot. And most teams have made peace with it because they thought it was unavoidable. It isn't anymore.
08:41If you're heading to Risk and Compliance and you want to see what testing your full population
08:45or any portion of it actually looks like in practice, come find us. Population testing,
08:51it's a natural extension of the platform our clients already trust. And we're looking forward
08:56to walking people through it in person at Risk and Compliance Conference, along with the many other
09:00fall conferences we have coming up. Phil, thank you so much for joining me today. Congratulations
09:06again on the launch of ACEs population testing. And we can't wait to see you at the MBA Risk and
09:14Compliance Conference. Very good. Thank you so much, Alison. Always enjoy talking with you.
