AI Transparency

How our AI systems work and how we ensure responsible use.

At EganForge LLC, we believe in transparency about how artificial intelligence powers our products and operations. This page explains what AI models we use, how decisions are made, what oversight mechanisms are in place, and how data is handled.

1. AI Models We Use

Our products are powered by leading AI models including:

  • Anthropic Claude: Used for analysis, content generation, code review, and decision synthesis.
  • OpenAI GPT-4o: Used for parallel analysis and cross-validation of AI outputs.
  • xAI Grok: Used as an additional perspective in our multi-model consensus system.

We continuously evaluate AI models and may add or replace providers as the technology landscape evolves.

2. Multi-Model Consensus

For critical decisions — including trading signals, product recommendations, and strategic analysis — we do not rely on a single AI model. Instead, we use a multi-model consensus approach:

  • Multiple AI models analyze the same question independently and in parallel.
  • A judge model synthesizes the responses, evaluating areas of agreement and disagreement.
  • The final output includes a confidence score based on model agreement (1, 2, or 3 models agreeing).
  • Low-agreement decisions are flagged for additional review or human oversight.

This approach reduces the risk of errors from any single model and improves the reliability of AI-generated outputs.

3. Human Oversight

EganForge operates a 5-tier governance model that determines when AI can act autonomously and when human approval is required:

  • Tier 1 — Routine Operations: Fully autonomous. Includes monitoring, data collection, and standard analysis. No human intervention needed.
  • Tier 2 — Standard Decisions: Autonomous with audit logging. Includes minor parameter adjustments and routine optimizations.
  • Tier 3 — Significant Decisions: Autonomous with multi-model consensus required. Includes trading decisions and product changes.
  • Tier 4 — High-Impact Decisions: Requires both multi-model consensus and human notification. Includes large capital allocations and strategic pivots.
  • Tier 5 — Critical / Emergency: Requires explicit human (CEO) approval. Includes system shutdowns, major financial decisions, and policy changes.

All decisions at every tier are logged with full audit trails including the decision rationale, model outputs, and governance tier applied.

4. AI Decision-Making in Trading

Our trading products use AI to analyze market conditions and generate trading signals. It is important to understand:

  • AI trading decisions are based on pattern recognition and statistical analysis of market data, not guaranteed predictions.
  • All trades are subject to risk management controls including position size limits, stop-loss thresholds, and portfolio concentration caps.
  • Trading AI can be paused or halted through our governance system.
  • Historical performance is logged and analyzed for continuous improvement, but past results do not guarantee future outcomes.

5. Data Handling for AI Processing

  • User inputs to our SaaS tools are sent to AI model providers strictly for generating responses. We do not use your data to train AI models.
  • Market data processed by our trading systems is publicly available financial data and does not contain personal information.
  • AI interaction logs may be retained for quality assurance but are anonymized and never shared with third parties for their training purposes.
  • You can request deletion of your AI interaction history by contacting privacy@eganforge.com.

6. Limitations of AI

We are transparent about what AI can and cannot do:

  • AI can make mistakes. Despite multi-model consensus, AI outputs may contain errors, hallucinations, or inaccuracies.
  • AI is not a substitute for professional advice. Our tools do not replace financial advisors, attorneys, or other qualified professionals.
  • AI reflects its training data. Models may exhibit biases present in their training data. We work to identify and mitigate these biases.
  • AI performance varies. Quality of outputs depends on the specificity and clarity of inputs, and performance may vary across different use cases.

7. Continuous Improvement

Our systems include built-in improvement mechanisms:

  • Automated performance tracking and outcome analysis.
  • Regular reviews of AI model accuracy and reliability.
  • Feedback loops that learn from both successes and failures.
  • Health monitoring with automatic issue detection and remediation.

8. Responsible AI Commitment

EganForge is committed to responsible AI use. We pledge to:

  • Maintain transparency about our AI systems and their capabilities.
  • Implement appropriate safeguards and human oversight.
  • Protect user data and privacy in all AI processing.
  • Continuously improve our systems based on real-world performance.
  • Comply with evolving AI regulations and best practices.

9. Questions

If you have questions about our AI systems or this transparency statement, please contact us: