Book Demo
Privacy policy

We cannot see your code. Period.

ByteBell serves a pre-indexed graph of your codebase to your AI tools over MCP. Your raw source is not stored, not read by us, and architecturally not available to us — which is why it cannot be used for anything, training included.

Last updated March 15, 2026

We do not use your data to train or distil AI models

ByteBell does not use customer data to retrain, fine-tune or distil AI models — because we physically cannot. We never see your code, your queries, or the AI responses generated by your coding tools. The MCP protocol means your AI tool (Cursor, Claude Code, Windsurf and the rest) talks to ByteBell's knowledge graph directly. We serve pre-indexed context; we do not intercept, log or store the conversation between you and your assistant.

What we store on our servers

The only information held on ByteBell's servers is:

  • MCP tokens — authentication tokens that let your AI tools connect to your ByteBell knowledge graph.
  • Usage metrics — how many MCP queries your team makes, for billing and rate limiting.
  • Server logs — standard application logs, used only to monitor for bugs, errors and service health.

That is it. No code. No queries. No AI conversations. No personal browsing data.

Knowledge graph data

When you connect your repositories, ByteBell builds a knowledge graph that captures structural relationships — dependency maps, API contract connections, service relationships and file-level metadata. That graph is:

  • Stored in your isolated environment (cloud-hosted, your VPC, or on-premises, depending on your plan).
  • Accessible only through your authenticated MCP tokens.
  • Deletable at any time — disconnect your repositories and the graph is destroyed.

Account information

When you create a ByteBell account, we collect:

  • Email address
  • Name
  • Company or organisation name

Used solely for account management, billing and support communication.

Data sharing

We do not:

  • Sell your data to third parties.
  • Use your data to train, retrain or distil any AI model.
  • Share your code, queries or usage patterns with anyone.
  • Use your data for advertising or marketing.
  • Give any third party access to your knowledge graph.

Security

  • All data encrypted in transit (TLS 1.3) and at rest (AES-256).
  • MCP token authentication on all API access.
  • Permission inheritance — ByteBell respects your existing GitHub and GitLab repository-level permissions.
  • Isolated tenant environments; your data is never co-mingled with another customer's.
  • On-premises and VPC deployment options for maximum control.

Your rights

You can:

  • Disconnect repositories and delete your knowledge graph at any time.
  • Delete your account and all associated data.
  • Export your usage data.
  • Request a full data audit by emailing saurav@bytebell.ai.

Contact

For privacy concerns: saurav@bytebell.ai.

Changes to this policy

We may update this policy. Changes are posted here with a new "last updated" date.