← Home
← Blog
Methodology

How PreCap Logic Works: The Methodology Behind the Assessment

·4 min read

Most tools in the small business lending space operate as black boxes. You enter information, a number appears, and you are expected to trust it. There is no explanation of how the result was produced, what inputs drove the output, or what would need to change to get a different answer.

PreCap Logic was built on a different premise: if a tool is going to tell a business owner where their file stands, it should be able to explain exactly why.

This page documents how the assessment works — the logic behind it, what it evaluates, and what the output is designed to mean.

Why a Deterministic Engine, Not an AI Chatbot

PreCap Logic's core engine is deterministic. That means the same inputs produce the same outputs every time, without variation, without randomness, and without a language model inventing an answer.

That design choice was deliberate. When the output of a tool affects how a business owner thinks about their readiness for debt, consistency and explainability matter. An AI chatbot might produce a confident-sounding answer that shifts subtly between sessions. A rules-based engine produces the same result for the same file, every time — and every factor that shaped the result is traceable.

AI is used in one specific, bounded role: as a translation layer. After the deterministic engine assigns a readiness classification, risk flags, and capital path, an AI model generates a plain-English summary of those outputs. It does not alter them. It does not override the engine's logic. It explains what the engine already decided.

The distinction matters: logic decides, AI explains.

The Five-Layer Evaluation Model

Each assessment runs through five sequential evaluation layers. Every layer adds signal to the picture before the final output is produced.

Layer 1 — Repayment foundation

The engine first asks how loan repayment would realistically be supported given the current file. This could come from business revenue, contract-based income, personal income streams, asset-backed support, or a combination. The source and stability of repayment capacity is the starting point for every other evaluation.

Layer 2 — Proof quality

Repayment capacity only matters if it can be documented. This layer evaluates what evidence currently exists — bank statements, deposit history, invoices, signed contracts, tax returns — and how well that evidence supports the story the file is trying to tell. A business with strong revenue but weak documentation carries a weaker file than its fundamentals suggest.

Layer 3 — Risk and pressure signals

The engine scans for patterns that typically weaken a file in a lender's evaluation: overdraft history, credit pressure, income volatility, missing documentation, unresolved derogatory marks, or structural issues that complicate underwriting. These signals do not automatically disqualify a file — but they are identified, weighted, and surfaced in the output.

Layer 4 — Strength signals

Simultaneously, the engine captures positive signals: consistent deposit patterns, clean account management, collateral availability, signed customer commitments, or a stronger credit profile. Strength signals affect the capital path output and are surfaced alongside risk flags so the user understands both sides of their current picture.

Layer 5 — Path alignment

The final layer matches the overall profile — repayment source, proof quality, risk signals, strengths, business stage, and scenario context — to the capital path that appears most realistic given the current file. This is not a lending decision. It is a readiness-oriented routing based on how the file looks today.

How Capital Paths Are Assigned

Capital paths are the output of the full five-layer evaluation. They represent where a business's file currently aligns, not a prediction of what a lender will decide.

The possible paths reflect the real range of small business capital situations:

Each path comes with a specific explanation of why the file landed there, which factors drove the result, and what the highest-priority next steps are. The goal is not to deliver a verdict — it is to make the next right move clear.

What the Output Means

Every PreCap Logic result is designed to answer four practical questions:

Where does the file currently stand? A readiness classification and score reflect the current strength of the file — not approval odds, not a credit score, not a lender's decision.

Why did it land there? The result surfaces the specific factors — both blockers and strengths — that produced the output. Nothing is hidden.

Which capital path is most realistic right now? The recommended path reflects where the file fits today, given its actual documentation, repayment structure, and risk profile.

What should change next? The action plan identifies the highest-leverage improvements — whether that is organizing documentation, stabilizing cash flow proof, resolving a credit issue, or clarifying use of proceeds.

The output is not a loan approval. It is a structured readiness picture designed to reduce confusion before a business owner enters a lending conversation.

What PreCap Logic Is Not

Transparency about what a tool does requires equal clarity about what it does not do.

PreCap Logic is not a lender, a loan broker, or a marketplace. It does not originate loans, refer users to specific lenders for compensation, or generate lending offers. It does not pull credit reports or access any financial data beyond what the user provides in the intake. It does not guarantee any funding outcome.

It is a preparation and readiness tool. Its purpose is to make pre-application preparation clearer, more consistent, and more actionable — for business owners navigating the capital system on their own, and for advisors helping clients build stronger files before the first lender conversation.


Ready to see where your business stands? Try PreCap Logic free at getprecap.com — no signup required.

See how ready your business is — in under 5 minutes.

Try PreCap Logic — Free →