Deterministic LLM Scoring With Fixed-Corpus Evidence Traceability

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current investment evaluation systems rely heavily on manual processes, are labor-intensive, lack scalability, and suffer from inconsistent and non-transparent ESG ratings, leading to inefficiencies and reduced investor confidence.

Innovation Solution

A system and method for deterministically generating reproducible evaluative scores using a large-language model (LLM) under a structured prompt framework, with a fixed, version-controlled corpus and multi-pass deterministic analysis, ensuring transparency and reproducibility through provenance metadata.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual assessment methods are used for investment analysis, then detailed and comprehensive evaluation can be achieved, but the process becomes labor-intensive and time-consuming

Engineering Contradiction:

Improveevaluation comprehensivenessVSAvoidanalysis efficiency

Core Design Contradiction:

Measurement precisionVSProductivity

Solution Approach 1:

The patent replaces manual mechanical assessment processes with an automated AI-based system that uses large language models to evaluate investment securities. The system automatically extracts information from financial statements, ESG disclosures, and regulatory filings, eliminating the need for manual review while maintaining comprehensive evaluation coverage.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system enables self-service evaluation where the AI model independently performs the entire analysis process without human intervention. The automated system extracts, analyzes, and scores investments based on predefined criteria, allowing the system to serve itself rather than requiring continuous manual oversight.

Inventive Principle:

Principle #25Self-service

2Measurement precision

If conventional manual analysis is performed, then in-depth assessment of limited securities is possible, but scalability to large universes of securities is restricted

Engineering Contradiction:

Improveassessment depthVSAvoidscalability

Core Design Contradiction:

Measurement precisionVSAdaptability or versatility

Solution Approach 1:

The patent replaces manual analysis mechanics with automated AI processing that can scale across large universes of securities. The system processes thousands of securities simultaneously by automatically extracting and analyzing data from multiple sources, enabling both in-depth assessment and broad scalability without trade-offs.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Quantity of substance

If ESG ratings are based on self-reported corporate data, then data availability is improved, but reliability and credibility are reduced due to potential bias and greenwashing

Engineering Contradiction:

Improvedata availabilityVSAvoidrating credibility

Core Design Contradiction:

Quantity of substanceVSReliability

Solution Approach 1:

The patent merges multiple data sources including self-reported corporate data, third-party verification, alternative data streams, and regulatory filings into a comprehensive evaluation system. This combination allows the system to cross-validate information and reduce bias while maintaining data availability from multiple perspectives.

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

The system incorporates feedback mechanisms where AI models continuously verify and cross-validate corporate disclosures against external data sources and regulatory requirements. This feedback loop detects inconsistencies and potential greenwashing claims, adjusting ratings based on verified information rather than relying solely on self-reported data.

Inventive Principle:

Principle #23Feedback

4Adaptability or versatility

If proprietary scoring models are used by rating agencies, then competitive advantage is maintained, but transparency and reproducibility are reduced

Engineering Contradiction:

Improvemodel flexibilityVSAvoidmethodology transparency

Core Design Contradiction:

Adaptability or versatilityVSLoss of information

Solution Approach 1:

The patent introduces an intermediary layer of standardized, transparent scoring criteria that mediates between the flexibility needed for competitive modeling and the transparency required for reproducibility. The system uses publicly documented rubrics and weighted criteria that allow full disclosure of methodology while maintaining the flexibility of AI-based analysis.

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS20260073450A1System and method for deterministically generating reproducible evaluative scores for a subject of analysis

Publication Date: 2026.03.12 VIALAB

AI summary

The present invention relates to a system and method for deterministically generating reproducible evaluative scores for a subject of analysis (e.g., a security). The system comprises a processor and memory storing instructions to: receive verified data describing the subject; store this data in a fixed and version-controlled corpus to define a static analytical context; execute a large-language model (LLM) under a structured prompt framework that directs a controlled scratch-pad reasoning process for preliminary interpretations and evidence extraction; perform a multi-pass deterministic analysis of the fixed corpus to produce structured, synthesized statements as reproducible evidentiary outputs; and finally, apply a rubric-based scoring engine that converts these statements into calibrated alignment scores and aggregates them to generate a composite deterministic score. This architecture ensures reproducibility, transparency, and auditability by anchoring the flexible analysis of the LLM and the final scoring logic to a known, unchanging evidence corpus.