Kamarin Lee

Independent Researcher

I build data and AI systems: data engineering, architecture, machine learning, and the governance layers underneath them.

Over the past decade I have delivered data and AI programs for Meta, T-Mobile, and other Fortune 100 organizations, and led data and machine learning work at UiPath and at a private-markets investment platform. My applied work in media-mix modeling is featured in Wiley’s Data Science For Dummies.

Working on a private-markets investment platform, I was close to how commercial real estate and private equity offerings are diligenced and reported. What stayed with me is how much of the record is assembled by hand: a track record arrives as a spreadsheet, terms live in a PDF, and the diligence that checks them is a series of phone calls.

I now work on what it should take for a number to be trusted when the records behind it are hard to reach. In 2026 I registered a study on public regulatory filings; the paper follows.

ORCID//SSRN GitHub LinkedIn//Research//Essays

Research

Verification and provenance in financial data
A Verification Standard for Filing Data
In progress
Provenance, per-field status, confidence labeling, versioned processing, published error rates.
Study on public regulatory filings
Registered 2026
Pre-registered and under embargo. Publication to follow.

Writing

Essays on verification
What “Verified” Should Mean
Aug 2026
The word does real economic work with no definition behind it.
Why Provenance Beats Freshness
Forthcoming
Pre-Registration Outside Academia
Forthcoming

Kamarin Lee

Brandeis University  |  B.F.A.

Bio

Kamarin Lee is an independent researcher working on verification and provenance in financial data: what it should take for a number to be trusted when the records behind it are hard to reach. His current work formalizes a verification standard for regulatory filing data and tests it empirically on public records.

His work spans two long-running threads. The first is applied measurement, beginning in marketing and consumer analytics and continuing through machine learning and media-mix modeling, work featured in Wiley’s Data Science For Dummies. The second is large-scale data and AI systems architecture, spanning private-markets investment infrastructure, enterprise automation at UiPath, and data and AI programs for Meta, T-Mobile, and other Fortune 100 organizations.

Research

Research interests: Verification and provenance in financial data. Measurement under partial and self-reported records. Identity resolution across regulatory identifiers. Point-in-time correctness.
Past: Applied machine learning and media-mix modeling. Enterprise data architecture.

A Verification Standard for Filing Data
In progress
A working specification: provenance to primary sources, per-field status, confidence labeling, versioned processing, and published error rates. Written to be implementable and auditable rather than aspirational.
Study on public regulatory filings
Registered 2026
Pre-registered study design, hypotheses, verification protocol, and analysis plan, timestamped before any study-window data was examined. Under embargo; publication to follow.

Selected Work

Applied work featured in Data Science For Dummies, Wiley
3rd ed.
Media-mix modeling in commercial practice.
Verification pipeline for regulatory filings
2026
Provenance-first ingestion with per-field status, confidence, and parser versioning; hand-verified core corpus with measured inter-verifier agreement.

Professional Activities

Principal architect, enterprise data and AI
2023–
Data and AI programs for Meta, T-Mobile, and other Fortune 100 organizations across financial services, healthcare, telecommunications, and consumer sectors.
Data and AI engineering — UiPath; private-markets investment platform
2019–2023
Enterprise automation platforms; investor and offering data systems for a private-markets investment platform covering commercial real estate and private equity vehicles.
Analytics and machine learning, media and consumer
2013–2019
Measurement, modeling, and data platform work across publishing, fashion, and consumer brands.

Contact

ORCID 0009-0000-0895-2450LinkedInGitHubSSRN
A full curriculum vitae, including complete engagement history, is available on request.