Real-Time Market Intel & Web Search

The Insights engine is no longer limited to its training data. We have activated a dedicated Web Search capability that allows the AI to browse the live internet for the latest market news, sentiment, and data points.

When you ask about "latest news on BTC" or current market events, the system now performs a real-time Google search, synthesizes key facts, and provides answer with citied sources. This ensures your trading decisions are backed by the most up-to-the-minute information available, not just historical data.

Smarter Insights with Knowledge Base RAG

Alongside the web, the Insights engine has received a major upgrade with the implementation of a Retrieval-Augmented Generation (RAG) pipeline for internal knowledge.

The system now actively references the platform's own authorized documentation. When you ask a question about platform mechanics or strategy, the AI retrieves relevant sections—from trading basics to advanced features—to provide answers that are technically accurate and specific to VTX Macro's unique environment.

Execution Context Transparency

We are demystifying the "Black Box" of AI trading. The system now captures a snapshot of the Execution Context at the exact moment of every trade decision.

This feature allows users to view the sanitized prompt templates and active settings that led to a specific Buy, Sell, or Hold call. By exposing the logic ecosystem—including the exact decision constraints and behavioral guidelines active during the trade—users can better understand, trust, and debug the AI's reasoning.

Unified Documentation System

We have consolidated the platform's information architecture. The user-facing Help page and the AI's internal Knowledge Base now share the same source material.

This unification ensures a "Single Source of Truth." Any update to the documentation immediately benefits both the human reader (via the Help page) and the AI Assistant, ensuring that the AI's understanding of the platform is always in sync with the user manual.

Performance & Core Improvements

• Gemini 3 Flash: The Insights extraction model has been switched to gemini-3-flash-preview. This upgrade provides faster response times and improved cost-efficiency for analyzing large amounts of market context.
• Configuration Refactor: System models have been modularized, making the platform easier to update as new models satisfy the platform's evolving needs.
• Secure Visibility: Enhanced sanitization logic ensures that while Execution Contexts are transparent, they remain secure.