Privacy Policy
1. Architectural Principle: Minimal Data Retention
E-NODE AI provides autonomous financial statement parsing, reconciliation, and audit analytics. Because our platform is purpose-built to analyze sensitive corporate accounting records, our system is designed to keep as little customer data as possible, for as short a time as possible.
Data Handling
Uploaded source files are processed in memory and not retained. Extracted figures are stored in a per-organisation workspace so you can review past runs, and you can view or request deletion of them at any time.
2. File Processing
When you submit financial filings or corporate ledgers for analysis:
- • Temporary Staging: Uploaded files are staged in RAM-backed temporary storage where available and deleted after analysis.
- • Response Caching: To avoid repeat processing, AI responses may be cached on our server.
- • Extracted Figures: Figures extracted from your files are kept in your organisation's workspace until you ask for them to be deleted.
3. Pre-Processing Identification Masking
Prior to algorithmic interpretation or extraction by language models, source documents pass through an automated privacy redaction layer.
Any detected personal identifiable attributes (such as individual signatory names, individual identification numbers, personal contact numbers, and personal bank account details) are masked or redacted in memory before statement line items enter downstream extraction pipelines.
4. Zero Model Training on Customer Data
We strictly forbid the use of customer financial materials for model training:
- • Customer documents, trial balance rows, and balance sheet reconciliation results are never utilized to train, retrain, fine-tune, or calibrate public or foundational artificial intelligence models.
5. Transmission Security
This website is served over HTTPS.
6. Early Access Inquiries & Contact
For inquiries regarding our technical verification architecture, cryptographic safeguards, or private beta onboarding protocols, please contact the founder directly: