By using RTIO, you benefit from an architecture designed around trust and precision:

• Zero Data Exposure to LLMs: Only schemas, column types, null ratios, and cardinality profiles are transmitted to the AI planning layer. Your raw data rows never leave the deterministic execution engine.
• Deterministic Execution: Every pipeline output is mathematically reproducible. The same input and configuration always produce the exact same result. Probabilistic variance is engineered out of the execution path.
• Human-in-the-Loop: The AI suggests cleaning plans; you approve every operation before execution. Nothing mutates your data without your explicit consent.
• Full Audit Trail: Every transformation is logged with SHA-256 hashes, tracking what changed, when, and via which operation — giving your compliance team complete lineage.

RTIO provides data preprocessing through multiple channels:

• Web Application: A visual node-based canvas where you build, inspect, and execute cleaning pipelines with real-time streaming progress.
• AI Chat: A conversational interface that profiles your data, suggests cleaning operations, and builds pipelines from natural language instructions.
• MCP Server: A metadata-only Model Context Protocol server for integration with AI agents and LLM workflows (coming soon).
• Workflow Automation: n8n nodes for embedding RTIO pipelines into scheduled or event-driven workflows (coming soon).
• Python SDK & CLI: Programmatic access for scripting, CI/CD integration, and batch processing (coming soon).

Supported formats include CSV, TSV, JSON, JSONL, Excel (.xlsx/.xls), and Apache Parquet. All data is converted to compressed Parquet upon ingestion for optimal performance.

RTIO is built for teams and organizations that need reliable, repeatable data preprocessing:

• Data Engineers building robust, auditable preprocessing pipelines.
• AI/ML Teams establishing clean training datasets with full provenance.
• Data Analysts standardizing and cleaning data without writing code.
• Regulated Organizations (Healthcare, Finance, Government) requiring deterministic execution with complete audit trails.

• You own your data. All datasets you upload, pipelines you create, and outputs you export remain your intellectual property.
• No training on your data. We do not use your uploaded datasets to train, fine-tune, or improve any AI or machine learning models.
• Export anytime. You can download your data and pipeline configurations at any time in standard formats.
• Deletion on request. When you delete your account or datasets, we remove them from our systems, subject to legal retention obligations.

As a user of the RTIO platform, you agree to:

• Review and approve the pipeline steps proposed by the AI before execution.
• Ensure you have the legal right to upload and process the data you submit to the platform.
• Comply with all applicable local, state, national, and international laws and regulations regarding data privacy and protection.
• Maintain the confidentiality of your account credentials and immediately notify RTIO of any unauthorized access.
• Not attempt to reverse-engineer, circumvent, or exploit the platform's security boundaries.

You may not use RTIO to:

• Process data that you do not have lawful authority to collect, store, or transform.
• Attempt to extract, reconstruct, or exfiltrate data belonging to other users or tenants.
• Interfere with, disrupt, or place undue burden on the platform or its infrastructure.
• Use the platform for automated decision-making that directly affects individuals without appropriate human oversight.
• Resell, sublicense, or redistribute the platform without a written agreement.

RTIO is architected for regulated industries. We implement:

• Multi-tenant data isolation with Row-Level Security (RLS).
• AES-256 encryption at rest and TLS 1.3 encryption in transit.
• Role-Based Access Control (RBAC) with workspace-level permissions.
• Zero-PII exposure to external APIs — the AI planning layer receives only statistical metadata.
• Comprehensive audit logging of all pipeline executions and data operations.

"As Is" Service: RTIO is provided on an "AS IS" and "AS AVAILABLE" basis. While we engineer for absolute precision, we make no warranties of any kind, whether express or implied, including but not limited to merchantability, fitness for a particular purpose, or non-infringement.

AI Planning vs. Deterministic Execution: The deterministic engine (Polars/DuckDB) guarantees reproducible outputs. The AI planning layer generates pipeline proposals based on metadata analysis. You are responsible for reviewing and approving all AI-suggested operations before execution.

Limitation of Liability: To the maximum extent permitted by law, RTIO shall not be liable for any indirect, incidental, special, consequential, or punitive damages arising from your use of the platform.

We may update these Terms from time to time. When we do, we will revise the "Last Updated" date and, for material changes, notify you via email or an in-app notification. Continued use of the platform after changes constitutes acceptance of the updated Terms.