AI Transparency & Ethics

Responsible AI that you can trust. Every decision explained, every outcome accountable, every process transparent.

Our AI Principles

At Jandojegs, we believe AI should augment human intelligence, not replace human judgment. Our platform is built on five foundational principles that ensure fairness, transparency, and accountability in every logistics decision.

1. Bias Mitigation

Ensuring fair and equitable AI decisions across all demographics, regions, and business sizes.

Diverse Training Data: Vendor selection models trained on carriers from 50+ countries, all size categories, and diverse ownership structures

Regular Fairness Audits: Quarterly audits across demographics, regions, and company sizes with public reporting

Human Review Trigger: Decisions exceeding $50K automatically flagged for human review to prevent algorithmic bias

2. Explainable AI (XAI)

Every AI decision comes with a clear explanation of why it was made.

"Why This Recommendation?": Every AI suggestion includes a plain-English explanation of the key factors

Feature Importance Scores: See which data points (price, reliability, speed) influenced each decision most

Confidence Intervals: All forecasts include upper/lower bounds with 95% confidence levels

3. Human-in-the-Loop Oversight

Critical decisions always require human approval. AI assists, humans decide.

High-Value RFQs: Quotes exceeding $100K require human approval before execution

Override Capability: Users can always override AI suggestions with documented reasoning

Feedback Loop: Users can mark AI errors for immediate model retraining and improvement

4. Model Transparency

Full disclosure of how our AI works, what data we use, and how models are trained.

Version Logging: Every prediction tagged with model version for full audit trail

Training Data Sources: Public disclosure of data sources used for model training

Performance Metrics: Monthly accuracy reports published on status page

5. Privacy-First AI

Your data is yours. We never use client data to train models without explicit consent.

Federated Learning: Models trained locally on your data, only aggregated insights shared

Differential Privacy: Statistical noise added to aggregate insights to protect individual data points

No Cross-Client Sharing: Your competitive data never used to benefit competitors

Accountability & Governance

Our AI decisions are reviewed quarterly by an independent Ethics Board composed of logistics experts, data scientists, and ethicists.

We publish an annual AI Transparency Report detailing model performance, bias audits, and user feedback trends.

Report AI-related concerns to: ethics@jandojegs.com

Questions About Our AI Ethics?

We're happy to discuss our AI principles, model architecture, and governance practices