AI Transparency

Our commitment

TranX AI is built on the principle of transparency over magic. We believe you should always understand how AI is being used to analyze your data, what models are involved, and how decisions are made. This page explains how AI works within our platform so you can use it with confidence.

How AI is used in TranX

TranX AI uses large language models (LLMs) to power the following capabilities:

• Natural language to SQL: Your plain-English questions are translated into SQL queries that run against your connected data sources.
• Data cleaning & processing: AI helps clean, normalize, and transform your raw data so it's ready for accurate analysis.
• Insight generation: AI analyzes query results to surface patterns, anomalies, and actionable recommendations.
• Narrative summaries: AI generates human-readable stories that explain the meaning behind your numbers.
• Dashboard generation: AI creates visualizations and dashboard layouts based on your data and questions.
• Automated monitoring & scheduling: AI powers scheduled tasks that automatically analyze and monitor your data, so you get timely alerts and reports without manual effort.

AI models we use

TranX AI currently uses large language models from Anthropic to power all AI capabilities across the platform. We continuously evaluate other models to provide users with the most reliable and accurate service possible. If we adopt additional model providers in the future, we will notify users in advance through this page and via in-app communications.

How your data is handled

• Data minimization: Only the schema metadata and relevant query results needed to answer your question are sent to AI models. We do not send your entire database.
• No training on your data: Your data is not used to train or fine-tune any AI models. We use API access with data-use opt-out agreements with our model providers.
• Encryption in transit: All data sent to AI model providers is encrypted using TLS.
• No persistent storage by providers: Our agreements with model providers ensure that your data is not retained beyond the duration of the API request.

Verifiability — show your work

Our goal is to give you an AI analyst that works like a real human analyst you just hired — someone you can ask anything, dig into any detail, and trust to show the full picture behind the numbers. Every AI-generated answer in TranX shows the underlying SQL query, the raw data it analyzed, and the reasoning steps taken. You can inspect, edit, and re-run any query yourself. The more you ask, the more context TranX AI builds — so you can trust it the same way you'd trust a dedicated analyst on your team.

Limitations & accuracy

AI-generated insights are probabilistic, not deterministic. While we implement validation layers to check SQL correctness and result plausibility, AI outputs may occasionally contain errors. We recommend treating AI-generated insights as a starting point for analysis and verifying critical business decisions against your own expertise and data.

Human oversight

TranX AI is designed as a human-in-the-loop system. AI does not take actions on your behalf — it generates analyses that you review and act on. You maintain full control over your data connections, the questions you ask, and how you use the results.

Updates to this page

As our AI capabilities evolve, we will update this page to reflect changes in the models we use, how data is processed, and any new AI features. Material changes will be communicated to active users.

Questions?

If you have questions about how AI is used in TranX, please contact us at info@tranx.io