Our Story

Built from the inside
of the machine.

We didn't pivot into M&A advisory. Every phase of our history built toward this — the forensic discipline, the econometric foundation, the underwriting experience.

Founded 2018

Business Analysis LLC

Business Analysis was founded in 2018 as a data-driven decision support firm. The original mandate was straightforward: quantify the relationships between enterprises, their competitors, customers, and macroeconomic environments. We built econometric models — structural equation modeling, multilevel regression, machine learning APIs — for clients across Southeast Europe, partnering with PhD researchers and university faculty to deliver analysis that most consulting firms in our market couldn't.

That work sharpened a specific skill set: the ability to extract signal from complex, fragmented datasets — and make it actionable. We spent years learning how to resolve noisy data, model non-linear relationships, and translate statistical output into business decisions.

In parallel, the founder's career took a forensic turn. Over 1,600 forensic court reports in financial operations between 2018 and 2025 — investigating assets, inventory, procurement records, and transaction-level fraud across hundreds of businesses. This wasn't academic. It was adversarial: every finding had to withstand legal scrutiny. The pattern recognition developed through forensic work — identifying irregularities in financial behavior, detecting when a company's numbers don't match its operations — became the foundation for how we now evaluate seller readiness.

In 2025, a fixed-term engagement at a Boston-based acquisition finance firm added the final layer: end-to-end SBA 7(a) underwriting across a 34-loan pipeline. Credit analysis, cash flow modeling, DSCR evaluation, collateral structuring, deal memoranda — the full acquisition finance stack. This wasn't consulting. It was underwriting real transactions, identifying the precise financial conditions under which a deal closes or collapses.

Three capabilities converged: doctoral-level statistical modeling, forensic-grade pattern recognition, and hands-on acquisition finance. The question became inevitable — what if we applied all of this to the one problem the middle market has never solved: identifying which companies are most likely to sell, before anyone else knows?

That is what Business Analysis is today. We engineered a proprietary universe of 300,000 middle market companies, continuously scored for seller readiness using graph neural networks that quantify both financial and non-financial variables — ownership behavior, succession dynamics, reinvestment patterns, filing anomalies. The models were built by a team led by a PhD in Data Science. The investment logic was architected by a team led by a PhD in International Management with forensic and underwriting experience.

We didn't pivot into M&A advisory. Every phase of our history built toward this. The forensic discipline ensures rigor. The econometric foundation enables the modeling. The underwriting experience grounds every recommendation in deal-level reality. And the two Principal-led teams who lead every engagement are the same teams who built the system — from the first regression model in 2018 to the GNN architecture operating today.


The Platform

How the system works

Our graph neural network evaluates every company in our universe across a continuous set of financial and non-financial variables. Traditional screening tools look at what a company reports. Our model looks at behavioral patterns — what ownership actions, filing sequences, and operational signals predict a transition event before it becomes public.

Universe Size
300,000+ companies
EBITDA Range
$1M – $10M
Data Sources
500+ integrated databases
Delivery
10–20 targets per month
Signal Type
Pre-market, off-listing
Scoring
Acquisition likelihood 0–100
01
Thesis Formalized Your acquisition criteria translated into a structured framework that governs how the model evaluates every firm in our universe.
02
Unified Intelligence Layer Corporate registries, lending records, licensing authorities, and proprietary datasets — resolved into a single profile per firm.
03
GNN: Financial and Non-Financial Signals A custom graph neural network that quantifies variables traditional models ignore — ownership behavior, succession dynamics, filing patterns.
04
Evaluated and Delivered, Face to Face Every target reviewed individually by your Principal-led team — then presented to you with full strategic context.

The Team

Every engagement is led by the people who built the system.

Investment Logic & Advisory
PhD, International Management
Background in financial forensics, econometric modeling, and SBA acquisition finance underwriting. Over 1,600 forensic court reports across financial operations. Former engagement at a Boston-based acquisition finance firm underwriting SBA 7(a) loan pipelines.
Model Architecture & Data Science
PhD, Data Science / Machine Learning
Designed and built the graph neural network infrastructure underlying the platform. Responsible for data pipeline architecture, model training strategy, and the heterogeneous graph schema that scores 300,000 companies for seller readiness.
2018 Founded
300K+ Companies Monitored
1,600+ Forensic Reports
500+ Data Sources

Predictive seller identification.
Built for individual acquirers.

PE firms have operated with this advantage for years. We bring it to you — with the rigor, precision, and personal involvement that institutional investors expect.