Jonathan Tsarong‑Blomker

I studied chaos and non-linear systems before I studied business analytics, and I still think of what I do as the physics of business.

An operation is a system with forces acting on it — demand, capacity, variation, the people doing the work. Most of those forces matter far less than they appear to. The work is finding the few that actually govern the outcome, measuring them precisely, and changing those.

Analytics and Lean Six Sigma give that work its precision, and they run as a loop rather than in sequence. Analytics locates where the system is losing time, quality, or money. Lean Six Sigma's structured tools — process mapping, root cause analysis, the discipline of stating a hypothesis before chasing it — turn that signal into sharp, testable explanations of why. Analytics then tests each explanation against the data at full scale, keeping what holds and discarding what does not. Where the evidence points, Lean Six Sigma supplies the method for changing the process and proving the change held. The result is surgical: a specific intervention in a specific place, backed by evidence, rather than a broad recommendation to improve.

Not every problem is a process problem, and a good share of my work sits well outside Lean Six Sigma's scope. Predictive models that see demand, attrition, or capacity pressure coming months before it arrives. Prescriptive models that go a step further and recommend the best response — how many people to hire and in which month, how much inventory to hold against a given risk. Simulation that tests a change before anyone commits capital to it. These are the tools behind most of the results on my example work page, and they are where a physics training earns its keep.

Two disciplines, one practitioner

I am a practicing analytics leader and a certified Lean Six Sigma Master Black Belt, and those are rarer together than either is alone. Lean Six Sigma came out of an era when measurement was expensive, so its tools assume sampling. That assumption has quietly expired. Bring modern data modeling to the same process questions and the method gets considerably sharper — you see the whole distribution rather than an inference about it, and you can tell whether a change actually moved the process or merely moved a number.

Where I sit in the work

Part of that precision is knowing what kind of problem you have before trying to solve it. Sometimes the data are genuinely the limiting factor — they sit in source systems nobody has extracted, or exist in a form no one can analyze. Other times the data are abundant and nobody has settled which question would change a decision. Those need different fixes, and mistaking one for the other is an expensive error.

A good deal of what matters happens upstream of any single analysis: data strategy, analytics roadmaps, and planning that sequences what gets built and when. I am precise about my own scope there. Connecting to source system APIs and building extraction pipelines is specialist engineering work, and I bring in or work alongside someone who does it. What I do is everything downstream of the raw landing tables — modeling that data into production layers an analyst can use, and defining what belongs there in the first place. That handoff is frequently where projects stall, and being clear about it early saves months.

What I try to avoid is building a foundation without knowing what it is for. A well-engineered pipeline feeding a well-built dashboard that answers the wrong question is still the wrong answer.

On AI

The team I was building was early to using AI as the ordinary way work got produced — analyses that took weeks came back in days, with review and consistent metric definitions built around the speed so it could be trusted. I remain the human in the loop: if I cannot explain why a result is true without pointing at the tool that produced it, it does not reach you.

Background

I started out in Chicago buying media in real time — a quantitative, high-frequency problem before most industries had one. That work took me to India for a year to build a team from scratch, and eventually to running a 100-plus person office there from Chicago, travelling back regularly. Over two of those years I also built a marketing analytics team split across both locations. Building analytics capability across continents and time zones taught me most of what I know about why analytics functions succeed or fail, and very little of it is technical.

For nine years alongside that career I owned and operated a digital marketing firm, with staff in the United States and a team I managed in India — carrying client outcomes and payroll at the same time. It is a useful thing to have done before advising other people about how their operations should run.

Most recently I spent the better part of a decade inside a large diagnostics and medical device company, moving from senior analyst to director and standing up five analytics teams: laboratories, supply chain and warehousing, manufacturing, logistics, and people and workforce analytics. On the people side that has meant novel metrics built on the Gallup and Great Place to Work frameworks, bottom-up workforce models, and hiring models that say how many to recruit and when. On the operations side, inventory risk models that free working capital, discrete event simulations that test a workflow change before anyone commits to it, capacity forecasting for equipment and labor, and ROI analysis that decides whether a project deserves its capital at all.

I work with a small number of clients at a time, which means the person you talk to is the person doing the analysis. If analysis is not going to answer your question, I would rather tell you that early than bill you to find out.

Based in

Madison, Wisconsin

Experience

Analyst to Director;
US and India

Teams built

Seven across two
continents; six were
analytics functions

Education

MS Analytics, Villanova
Physics, UW–Madison
chaos & non-linear systems

Certification

Lean Six Sigma
Master Black Belt
Professional Scrum Master I
Scrum Product Owner I

Executive
education

Strategic Leadership
Building & Leading Highly
Effective Teams
Villanova University

Focus

People and workforce;
operations;
analytics strategy;
AI in the analytics function

Methods

Predictive and prescriptive modeling
Discrete event simulation
Lean / Six Sigma
Risk and ROI analysis

Availability

Taking new work now;
a few clients at a time