Please log in to continue
WalkMe stores your data in the region chosen during account setup.
We have data centers in the US and EU that follow local privacy and compliance laws. If you're not sure which region to select, check with your admin.
By default, WalkMe's AI features run on WalkMe-managed large language models (LLMs), currently powered by Azure OpenAI. This is the recommended setup for most organizations: it is fully optimized for WalkMe AI out-of-the-box, validated for performance, and requires no additional configuration.
If your organization has a strict internal compliance requirement to use a specific, approved LLM, you may be eligible for Custom LLM (previously referred to as BYOLLM). This is an exception, not the default, and is relevant only in specific, rare cases.
Is Custom LLM right for you?
Custom LLM is a custom configuration option, not a standard product SKU. It is designed for organizations with a mandatory internal requirement to work with a specific, pre-approved LLM.
It is worth noting that WalkMe does not train models on your data, and all processing uses secure protocols. If your concern is primarily around data privacy or residency, reviewing WalkMe's security and data handling documentation is a good first step, as WalkMe's managed LLM may already address those requirements.
If after reviewing the documentation your organization still has a compliance-driven need to use your own LLM, contact your Customer Success Manager to discuss whether Custom LLM is applicable to your situation.
Custom LLM uses a hybrid deployment model. In this setup, queries are routed to your LLM, which runs within your own environment. For AI Answers queries, the WalkMe crawler stores content within your environment and pulls from it before submitting queries to your LLM.
Once Custom LLM is configured, you can monitor your AI consumption through the AI Usage Dashboard. The dashboard gives you visibility into how WalkMe's AI features are being used across your account, including query volume and usage trends.
This is especially useful when running a Custom LLM configuration, as it helps you track consumption against your own model's capacity and costs.
For full details on the dashboard:
| WalkMe Managed LLM (Default) | Custom LLM | |
|---|---|---|
| Setup | Ready out of the box | Requires coordination with WalkMe's product team |
| Optimization | Fully optimized for WalkMe AI | Depends on your model; not guaranteed |
| Supported models | GPT-4o, selected GPT-5 models | Azure AI interfaces and models |
| Data privacy | Secure processing; WalkMe does not train on your data | Handled by your LLM provider's policies |
| Who it's for | Most organizations | Organizations with a mandatory compliance requirement to use a specific LLM |
| Risk | Low | Latency, timeouts, and degraded experience if the model is not validated |