About Four Sigmas
Most practitioners who call Four Sigmas are not calling because everything is going well. They are dealing with an audit that has gone sideways, a sample that is being challenged, or a methodology question they cannot answer confidently – and the stakes are high. That is exactly the situation Four Sigmas was built for.
How This Started
Ivan did not build Sampling Maestro as an academic exercise. He built it because practitioners kept calling him with problems the existing tools could not solve, and because nothing on the market was built for the specific demands of unclaimed property and indirect tax examinations.
Ivan Tasic spent over 15 years solving complex statistical problems across economic consulting, industry, and academia at PwC, Duff & Phelps, NERA Economic Consulting, and United Airlines, and before all of that, earning a Ph.D. in Economics from Texas A&M University.
He was first introduced to unclaimed property in 2012, when a colleague at Duff & Phelps pulled him into an audit support engagement. The statistical challenges were immediately interesting: highly skewed dollar-value populations, adversarial examination settings, and methodology decisions with real legal and financial consequences.
Years later – after moving on to other roles – the calls kept coming. Former colleagues would reach out with problems: a sample that needed to be reviewed, a methodology question with a deposition looming, an examiner’s approach that did not look right but was hard to articulate why. Ivan kept helping, because the problems were genuinely hard and genuinely mattered.
Eventually the calls became frequent enough that the answer was obvious. In May 2021, he founded Four Sigmas, LLC.
The Work
Four Sigmas specializes in statistical sampling for unclaimed property examinations and indirect tax audits – the kind of work where the methodology has to hold up under scrutiny, in litigation, and in front of regulators who know how to push back.
The core finding that drives everything Four Sigmas does: attribute sampling formulas are routinely misapplied to dollar-value populations in these examinations. This is not a subtle technical quibble. It is a mathematical mismatch between formula and data type that can produce sample sizes off by orders of magnitude and confidence intervals that lack their stated coverage probability. Ivan documented this problem rigorously, and then built the tool that fixes it.
That tool is Sampling Maestro.
Sampling Maestro
A purpose-built statistical sampling engine with three specialized modes: Transaction Sampling for classical stratified variable sampling, Dollar Sampling (MUS) for Poisson-based monetary unit sampling with Stringer bound evaluation, and Sample Review for forensic examination of third-party samples.
Built on Cochran, Neyman, and Stringer. Every calculation traces to a peer-reviewed citation. Every result is reproducible. Every report is documented to withstand a Daubert challenge.
Explore Sampling MaestroBackground
Ivan has mentored diverse teams, communicated complex statistical methodology to executives and regulators, and supported engagements that have withstood legal and regulatory scrutiny.
The Mission
Statistics can be intimidating – especially when the stakes are high, the timeline is short, and the opposing side has experts of their own. Ivan built Four Sigmas for exactly that situation: to be the person practitioners can call when they need a straight answer, rigorous methodology, and documentation that holds up.
If you are facing an examination, evaluating a sample, or simply trying to understand whether your current approach is defensible – reach out. That is what Four Sigmas is here for.
Request a Live Demo Get In Touch