Security & Compliance
We maintain compliance with major security standards and regularly undergo security audits.
SOC 2 Type II
Compliant with SOC 2 Type II security standards.
GDPR
Full compliance with GDPR data protection requirements.
Data Protection
Encrypted by default
All data is encrypted in transit using TLS 1.3 and at rest using AES-256. You can also bring your own encryption keys (BYOK) for full ownership.
You decide what gets stored
Helply processes ticket data in real time. By default, nothing is saved unless you configure it. Custom retention policies let you control what stays and what gets deleted.
Tight access controls
Role-based permissions and data masking ensure agents only see what they need. Full access logs give you visibility into every interaction.
Responsible Disclosure
We maintain an open Responsible Disclosure Program to encourage security researchers to report vulnerabilities. If you believe you've found a potential issue, we want to hear from you.
What sets Helply privacy and protection apart
Fast, secure setup. Helply plugs into your existing support stack - no dev time, no long onboarding.
Works with your stack. Helply integrates directly with platforms like Groove, Zendesk, and Freshdesk.
Custom control over data. You decide what Helply stores, what gets deleted, and who can see what.
Built-in audit visibility. Track who accessed what, when, and from where.
GDPR-compliant by design. Helply meets GDPR standards out of the box.
Security-first architecture. From sandboxed model responses to strict runtime protections, Helply is built from the ground up.
Helply Security & Data Protection FAQ.
Do you use customer support tickets to train the AI agent?
Do you track how much sensitive data is redacted from support tickets?
Are agent responses ever generated from raw ticket data?
Do you redact sensitive data from user prompts in real time?
Are there guardrails against prompt injection or prompt manipulation attacks?
Can users manipulate the AI agent or cause it to learn harmful behavior?
How does Helply handle hallucinations and bias?
Is there an audit log or training history available?
Do you have a SOC 2 certification?
Where is customer data stored?
Do you use customer data to train your base AI models?
Do third-party model providers (e.g., OpenAI, Anthropic) train their models on our data?
Is Helply GDPR compliant?
What subprocessors does Helply use?
Does Helply offer domain allowlisting?
What authentication options are available?
How is rate limiting handled?
Is data encrypted?
Does Helply have a bug bounty program?
How can we report a security concern?
We guarantee a 65% AI resolution rate in 90 days, or you pay nothing.
End-to-end support conversations resolved by an AI support agent that takes real actions, not just answers questions.