top of page
LatentSense_Soft_Blue_Background_Web.webp

Evidence-First Reasoning

ReX

Investigate the evidence. Defend the conclusion.

ReX combines deep numerical analysis with language-based, logical-semantic reasoning to investigate complex questions, examine contradictions, surface conflicts, and produce conclusions grounded in verifable evidence.

rxMaps provides the required semantic foundation for ReX advanced reasoning.

LatentSense_Soft_Blue_Background_Web.webp

ReX IN ACTION

​One question. Multiple sources.
An inspectable conclusion.

See how ReX combines numerical and documentary evidence, evaluates conflicting claims, and identifies what the evidence supports.

Investigation question

"Did the equipment operate within its permitted temperature limit?"

LatentSense_Soft_Blue_Background_Web.webp
EVIDENCE INSPECTOR & ARGUMENT VIEW

Inspect the evidence. Follow the reasoning.

ReX goes beyond presenting a conclusion. Explore the original source passages, examine supporting and conflicting evidence, and follow the reasoning paths connecting claims to their supporting information.

Inspect evidence directly in its original source context

Investigate supporting claims, contradictions, and governing requirements

Trace how evidence contributes to the final conclusion

ReX_Evidence_Inspector_Web.webp

Evidence Inspector

ReX_Argument_View_Web.webp

Argument View

A CLOSER LOOK

See ReX reasoning in action.

Watch how ReX investigates questions, connects conclusions to supporting evidence, and makes reasoning paths and original sources available for inspection.

Evidence-First AI Walk-through

CORE CAPABILITIES

From complex evidence to inspectable findings.

ReX combines advanced numerical and language-based reasoning with evidence investigation, provenance tracing, and audit-ready reporting—all built on the persistent semantic knowledge foundation provided by rxMaps.

Integrated Numerical & Language-Based Reasoning

Analyze measurements, trends, and telemetry alongside specifications, policies, and operational context to investigate complex questions.

Evidence Inspector & Provenance
 

Inspect source passages, evidence collections, provenance, and reasoning paths to understand how findings are supported.

Evidence-First AI Reasoning

Examine claims, relationships, and contradictions, connecting analytical findings to identifiable supporting evidence.

Evidence Reports & Audit Packages

Generate structured evidence reports and portable ZIP packages containing ReX evidence, the available rxMap manifest, and SHA-256 integrity information.

Evidence Collection & Mapping

Discover, isolate, and organize evidence relevant to an investigation, revealing relationships and supporting evidence chains.

Model-Agnostic Enterprise Integration

Extend existing LLMs, AI agents, and enterprise applications through REST APIs and the Python SDK without foundation-model retraining.

Government & Public Sector

LatentSense_Soft_Blue_Background_Web.webp

EVIDENCE COLLECTIONS & AUDIT PACKAGES

From investigation to audit-ready evidence.

ReX goes beyond inspectable reasoning by organizing evidence into collections, generating structured reports, and packaging supporting information for independent review. Preserve the evidence, claims, source references, and reasoning associated with an investigation in portable, integrity-verifiable outputs.

Evidence Collections

Organize related inquiries, claims, and reasoning steps in one navigable collection.

evidence-collection_edited.jpg

What it provides

Inquiry summaries and status

Claim and trace step counts

Direct access to argument and inspection view

OUTCOME: Investigations organized

Evidence Reports

Select inquiry responses and supporting evidence to generate a structured PDF report.

evidence_report_edited.jpg

What it provides

Selected inquiry responses

Supporting claims and source references

Report-ready investigation record

OUTCOME: Findings documented

Audit Packages

Choose which investigations and related artifacts to include in a portable ZIP export.

Audit_package_edited.jpg

What it provides

Inquiries, mapped evidence, reasoning

Source files and JSON exports

SHA-256 integrity

OUTCOME: Evidence made portable

LatentSense_Soft_Blue_Background_Web.webp

INTEGRATION

Add evidence-first reasoning to your existing AI Architecture

Use your existing agents and applications

Integrate through REST APIs or the Latent-Sense Python SDK. Extend existing LLM applications, AI Agents, analytics, and workflows - including Amazon Bedrock-based solutions without replacing your current stack.

Built on rxMaps for advanced reasoning

ReX performs evidence-first reasoning over persistent semantic knowledge foundation created by rxMaps, helping teams investigate questions across numerical and language evidence and trace findings back to their sources. Learn more about rxMaps.

ReX access and outputs

ENTERPRISE OUTCOMES

Complex evidence. More defensible decisions.

ReX helps organizations investigate complex questions, connect findings to supporting evidence, and produce conclusions that can be inspected and reviewed. Built on rxMaps, it brings numerical and language-based reasoning into existing enterprise decision workflows.

Accelerate Investigations

Identify and organize relevant evidence across interconnected records, measurements, and documents.

Improve Decision Intelligence

Examine complex numerical observations alongside requirements, claims, and operational context.

Support Defensible Conclusions

Connect findings to inspectable evidence, provenance, and reasoning paths.

Simplify Audit Preparation

Generate evidence reports and portable, integrity-verifiable audit artifacts.

PUTTING REX TO WORK

Extend your team's capabilities with evidence-first reasoning.

ReX brings advanced numerical and language-based reasoning to existing applications, AI agents, and operational workflows. Built on rxMaps, it enables teams to investigate complex questions, examine supporting evidence, and produce traceable findings without replacing their existing AI infrastructure.

AI Development & DevOps

Embed evidence-first reasoning into applications, development environments, and automated workflows through APIs and SDKs.

​

Support debugging, technical investigations, and traceability across AI-driven operations.

Observability & Operations Teams

Investigate anomalies by connecting telemetry, performance measurements, logs, and operational documentation.

​

Correlate events, examine potential causes, and inspect the evidence behind operational findings.

AI Agents & Workflow Automation Teams

Equip existing agents and work bots with advanced evidence-grounded reasoning.

 

Support complex investigations and automated workflows with inspectable findings, provenance, and audit outputs.

START WITH WHAT YOU WANT TO UNDERSTAND

Go beyond answers.
Gain deeper understanding.

​ReX reasons across your connected body of knowledge, bringing together numerical and language-based evidence to reveal relationships, evaluate claims, and explain what the evidence supports.

LST - Cognitive AI

Latent-Sense Technologies

112 -970 Burrard Street, 

Unit 1330, Vancouver,

BC V6Z 2R4, Canada.

​

  • LinkedIn

LinkedIn.

© 2026 by Latent-Sense Technologies Inc. All Rights Reserved.

bottom of page