The Security Practices page describes the administrative, technical, and physical controls applicable to AgentByte.ai.
Hosting & Architecture
AgentByte.ai is available as a cloud-based service.
Cloud-based (hosted) services
This infrastructure for AgentByte.ai is provided and hosted by Amazon Web Services, Inc. (“AWS”). Learn more on theAWS Security website. Details on SOC and certifications are on theAWS Compliance website.
In addition to AWS, AgentByte.ai uses vector database services from Pinecone Systems, Inc.
Our cloud-based version uses OpenAI's API for AI inference. See theOpenAI Enterprise Privacy page for their practices.
Storage of Customer Code
AgentByte.ai does not store customer code. We may store vector embeddings of file paths, documentation, and AI-generated docstrings, but code is pulled on an "as-needed" basis from the customer's host (e.g., GitHub) only with explicit consent or if publicly licensed.
For on-prem deployments, your code never leaves your servers or cloud.
For cloud deployments, any ephemeral code storage occurs on AWS services provisioned by AgentByte.ai.
Storage of Customer Data
We store chat logs in AWS DynamoDB. AgentByte.ai personnel may access logs to provide support. In self-hosted deployments, logs remain on customer infrastructure.
Confidentiality & Security Controls
Confidentiality
Strict controls govern employee access to Customer Data. Limited, audited access may be used to resolve support issues. Employees are bound by confidentiality obligations and our internal security policies.
Return & Deletion of Customer Data
Within 30 days post-termination, customers may request return of Customer Data (if not already deleted). Admins can delete Customer Data at any time; production deletion occurs within 24 hours, with backups destroyed within 30 days (timelines may pause during an active incident investigation).
Monitoring & Validation
Certificates
AgentByte.ai is SOC 2 Type II compliant. Request our report atsecurity@agentbyte.ai. We align with prevailing industry standards and successors.
Audits
We undergo internal and external security assessments, plus continuous automated scanning of our platform. External audit reports are available upon request atsecurity@agentbyte.ai.
Personnel
Background checks, onboarding security training, and ongoing education are required. All employees sign comprehensive information security policies.
Questions About Security?
Reach out at security@agentbyte.ai — we'll get right back to you.