“Use more AI,” says your boss, your execs, your LinkedIn feed, your coworkers, your favorite Reddit communities, and the list goes on…
At the same time, data privacy, compliance, IP protection, infrastructure governance, and cost oversight are more important than ever, so those requirements often also end up tragically halting your AI plans.
This is often because:
• Internal policy mandates pre-approved AI providers only
• Data residency requirements block public internet inference
• Shared AI infrastructure is not permissible under some compliance rules
• Security teams worry about IP leakage or sensitive data exposure
You are told to use more AI, but only if it fits perfectly within all these boundaries.
This is exactly why, in line with our commitment to choice, control, and flexibility, we are introducing Bring Your Own Model (BYOM) to Cribl AI.
The fix: BYOM
BYOM gives you the control to power Cribl’s AI features with your preferred AI backend while maintaining full control over your data privacy, residency, and compliance posture.
You can connect your preferred LLM to Cribl, choose the right model for each use case, and keep your own authentication and access controls, all while preserving the seamless native experience across AI powered capabilities like Cribl Guard background detection, natural language pipeline builder, Cribl Search investigations, and more.
This means you can unlock the automation and productivity of Cribl’s AI capabilities while staying aligned with governance standards and internal operational policies and getting more value from any existing enterprise AI investments.
You get the power of Cribl AI on your terms.
Of course, not every organization needs or wants to bring their own model. If you prefer a fully managed experience, our AI capabilities continue to work out-of-the-box using the default Cribl-managed AI backend with no additional setup required. BYOM simply adds another option for organizations that need greater control. Whether you choose to continue on the default backend or connect your own provider, the experience of Cribl AI features remains the same.
Who benefits?
BYOM is ideal for organizations where governance is not optional.
Regulated industries such as financial services, healthcare, government, and utilities can now enable AI within Cribl products without violating compliance mandates or data handling policies. If you operate in air gapped or internet restricted environments, BYOM allows AI to run within your approved architecture rather than forcing you into shared infrastructure that does not meet your policies.
It is also ideal for large enterprises that have already standardized on an AI stack and want to extend those investments into Cribl. Organizations with strict regulatory or AI governance policies gain clear visibility and control over model selection, resource optimization, and data handling.
Why it matters
Maintain data privacy and compliance
Use Cribl AI enabled features with your own pre-approved backend so you can isolate sensitive data, control exactly where inference runs, and ensure AI usage aligns with your internal governance standards. Whether your requirements are driven by regulatory mandates, contractual obligations, or internal policy, BYOM keeps model execution within your approved boundaries and removes the blockers tied to shared infrastructure.
Unified visibility over enterprise AI
Many organizations route AI traffic through internal gateways such as LiteLLM to monitor usage, enforce policies, manage keys, and track costs across the enterprise. With BYOM, Cribl can connect directly to these gateways instead of calling model providers on its own. This means all AI requests from Cribl flow through the same controls your organization already uses for governance, observability, and cost management, allowing you to extend your existing AI platform strategy into your Cribl deployment without adding new oversight gaps.
Have control and choice over the best model
Select the provider and model that best balance performance, cost, and accuracy for your specific use cases. Currently, you can choose the model that powers each Workspace. In future releases, you will also be able to select multiple models within a Workspace and assign different models to different AI features. Because different use cases require different tradeoffs, BYOM gives you the flexibility to make those decisions and optimize for what matters most each.
Leverage existing investments
Get more value from the AI platforms your organization has already standardized on. BYOM lets you extend those investments into Cribl, helping consolidate AI usage and strategy while maximizing the return on your enterprise AI spend.
Unlock full productivity
Accelerate investigations in Cribl Search, enhance background detection in Cribl Guard, build pipelines using natural language, and much more, all powered by your trusted AI backend. BYOM ensures teams gain the productivity benefits of AI driven workflows while staying within the architectural and governance guardrails your organization requires.
Enable AI in restricted environments
Bring AI capabilities to on-prem and highly regulated deployments without compromising data residency, privacy, or regulatory requirements. BYOM enables AI innovation for organizations operating in air-gapped or internet-restricted environments.
How to use BYOM in 4.17
To get started, upgrade to 4.17, ensure you have Owner or Admin permissions, and confirm that Cribl Copilot is enabled for your deployment. For more details, check out our clear and simple documentation.
Next, navigate to Settings → Global → AI Settings and click Use Custom AI Provider. From there, you can select the provider type you want to configure.
Enter your Provider ID, optional description, deployment URL, and API key. You can choose to test the connection to confirm everything is configured correctly before clicking Save.
Once saved, a provider card with details appears at the top of the page displaying and from that point forward new AI requests will use your configured provider for supported features.
Keep in mind for this initial release in 4.17, a single AI provider can be selected per Workspace. Future phases will introduce support for multiple providers with selection per feature, along with broader model catalogs, evaluation capabilities, and recommended model mappings.
The experience of Cribl AI features remains seamless with the same UI and workflows, now powered by your AI backend.
Conclusion
Your Cribl AI dreams (yes, we know telemetry management is all you dream about too) do not have to be crushed by compliance and governance policies.
With Bring Your Own Model, those dreams can become reality.
You get the automation and productivity of Cribl AI and more value from your existing AI stack investments while meeting your privacy, governance, and infrastructure requirements.
Cribl AI on your terms. Try it out in 4.17 today.