Data Privacy & AI Telemetry Policy
At Assembly Minds, we engineer AI systems that treat data security and privacy as fundamental architectural pillars. We strictly adhere to enterprise-grade compliance, zero-data retention on public models, and rigorous telemetry processing policies to safeguard all client and customer interactions.
01. Scope & AI Telemetry Ingestion
This policy outlines how Assembly Minds captures, processes, and protects data across all our AI service endpoints. Our ingest pipelines encompass inbound telephony audio streams via Twilio/Vonage, SMS and WhatsApp webhook payloads, and bi-directional CRM synchronizations with platforms such as HubSpot, Salesforce, and GoHighLevel.
We utilize real-time stream processing architecture with ephemeral buffers. Payload verification and sanitization occur instantaneously at the edge, ensuring that malformed or unauthorized telemetry is rejected before entering our core network. Telemetry ingested during active sessions is strictly constrained to the memory space required for inference execution.
02. Zero-Data Retention (ZDR) & LLM Sub-processors
Assembly Minds partners with industry-leading Foundation Model providers and Speech-to-Text inference engines. We operate under strict enterprise agreements with our sub-processors (including OpenAI, Anthropic, Deepgram, and Twilio).
Zero-Data Retention Guarantee
We explicitly guarantee that client confidential data, raw caller audio, and voice transcripts are never utilized to train, fine-tune, or improve public Large Language Models (LLMs). All enterprise zero-retention sub-processor covenants are actively monitored and programmatically enforced.
03. Data Storage & 256-bit Encryption Standards
Security is paramount at all layers of the Assembly Minds stack. Data at rest is secured using AES-256 encryption. All data in transit is protected via TLS 1.3 to prevent interception or tampering.
We implement end-to-end transport layer security across all telephony SIP trunks and internal API microservices. Cryptographic keys are managed via isolated KMS instances with automated key rotation policies in place, ensuring strict tenant isolation and preventing cross-client data contamination.
04. 30-Day Rolling Log Purge Policy
Our voice agents and AI architectures are designed around ephemeral processing. Telephony audio packets and transient memory states are immediately destroyed upon call conclusion or websocket termination.
For troubleshooting, quality assurance, and billing auditability, diagnostic and system execution logs are maintained temporarily. These logs are subject to an automated rolling schedule and are hard-deleted after 30 days. This ensures that potentially sensitive debug information is purged consistently without requiring manual intervention.
05. GDPR & CCPA Compliance and Data Subject Access Requests
Assembly Minds is fully committed to upholding international privacy frameworks, specifically Articles 15-22 of the General Data Protection Regulation (GDPR) and the California Consumer Privacy Act (CCPA) standards.
- Right to Know & Access: Users can request details regarding the specific pieces of personal information we have collected.
- Right to Rectify & Delete: We provide mechanisms for data subjects to correct inaccurate data or request complete erasure from our systems.
- Right to Opt-Out: We explicitly state that Assembly Minds does not and will not sell personal data.
All Data Subject Access Requests (DSARs) are processed and executed within the legally mandated 30-day turnaround period.
06. Data Protection Officer & Privacy Inquiries
For detailed inquiries regarding our AI ethics framework, data ingestion practices, sub-processor agreements, or to initiate a Data Subject Access Request, please contact our Data Protection Officer (DPO) and compliance team directly.
Direct Contact Email: contact@assemblyminds.com
In the event of a suspected vulnerability or security concern, please include "SECURITY ESCALATION" in your subject line to trigger our immediate incident response procedure.