Skip to content

Commit 62b1f63

Browse files
notgiorginotgiorgi
authored
docs: update AI Analyst environment variables and add Bedrock support (#261)
1 parent b9c673d commit 62b1f63

File tree

1 file changed

+57
-20
lines changed

1 file changed

+57
-20
lines changed

‎self-host/customize-deployment/environment-variables.mdx‎

Lines changed: 57 additions & 20 deletions
Original file line numberDiff line numberDiff line change
@@ -192,34 +192,42 @@ These variables allow you to configure [S3 Object Storage](/self-host/customize-
192192

193193
These variables enable you to configure the [AI Analyst functionality](/guides/ai-agents). Note that you will need an **Enterprise Licence Key** for this functionality.
194194

195-
| Variable | Description | Required? | Default |
196-
| ---------------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------- | --------- | -------- |
197-
| `AI_COPILOT_ENABLED` | Enables/Disables AI Analyst functionality | | `false` |
198-
| `ASK_AI_BUTTON_ENABLED` | Enables the "Ask AI" button in the interface for direct access to AI agents, when disabled agents can be acessed from `/ai-agents` route | | `false` |
199-
| `AI_EMBEDDING_ENABLED` | Enables AI embedding functionality for verified answers similarity matching | | `false` |
200-
| `AI_DEFAULT_PROVIDER` | Default AI provider to use (`openai`, `azure`, `anthropic`, `openrouter`) | | `openai` |
201-
| `AI_DEFAULT_EMBEDDING_PROVIDER` | Default AI provider for embeddings (currently only `openai` supported). Falls back to `AI_DEFAULT_PROVIDER` if not set | | `openai` |
202-
| `AI_COPILOT_DEBUG_LOGGING_ENABLED` | Enables debug logging for AI Copilot | | `false` |
203-
| `AI_COPILOT_TELEMETRY_ENABLED` | Enables telemetry for AI Copilot | | `false` |
204-
| `AI_COPILOT_REQUIRES_FEATURE_FLAG` | Requires a feature flag to use AI Copilot | | `false` |
195+
| Variable | Description | Required? | Default |
196+
| ---------------------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------- | --------- | -------- |
197+
| `AI_COPILOT_ENABLED` | Enables/Disables AI Analyst functionality | | `false` |
198+
| `ASK_AI_BUTTON_ENABLED` | Enables the "Ask AI" button in the interface for direct access to AI agents, when disabled agents can be acessed from `/ai-agents` route | | `false` |
199+
| `AI_EMBEDDING_ENABLED` | Enables AI embedding functionality for verified answers similarity matching | | `false` |
200+
| `AI_DEFAULT_PROVIDER` | Default AI provider to use (`openai`, `azure`, `anthropic`, `openrouter`, `bedrock`) | | `openai` |
201+
| `AI_DEFAULT_EMBEDDING_PROVIDER` | Default AI provider for embeddings (`openai`, `bedrock`) | | `openai` |
202+
| `AI_COPILOT_DEBUG_LOGGING_ENABLED` | Enables debug logging for AI Copilot | | `false` |
203+
| `AI_COPILOT_TELEMETRY_ENABLED` | Enables telemetry for AI Copilot | | `false` |
204+
| `AI_COPILOT_REQUIRES_FEATURE_FLAG` | Requires a feature flag to use AI Copilot | | `false` |
205+
| `AI_COPILOT_MAX_QUERY_LIMIT` | Maximum number of rows returned in AI-generated queries | | `500` |
206+
| `AI_VERIFIED_ANSWER_SIMILARITY_THRESHOLD`| Similarity threshold (0-1) for verified answer matching | | `0.6` |
205207

206208
The AI Analyst supports multiple providers for flexibility. Choose one of the provider configurations below based on your preferred AI service. **OpenAI integration is the recommended option as it is the most tested and stable implementation.**
207209

210+
#### Minimum Required Setup
211+
212+
To enable AI Analyst, set `AI_COPILOT_ENABLED=true` and provide an API key for `AI_DEFAULT_PROVIDER` (e.g., `OPENAI_API_KEY` for OpenAI, `ANTHROPIC_API_KEY` for Anthropic).
213+
208214
#### OpenAI Configuration
209215

210-
| Variable | Description | Required? | Default |
211-
| ---------------------------- | ------------------------------------------- | -------------------------- | ------------------------- |
212-
| `OPENAI_API_KEY` | API key for OpenAI | Required when using OpenAI | |
213-
| `OPENAI_MODEL_NAME` | OpenAI model name to use | | `gpt-4.1` |
214-
| `OPENAI_EMBEDDING_MODEL` | OpenAI embedding model for verified answers | | `text-embedding-3-small` |
215-
| `OPENAI_BASE_URL` | Optional base URL for OpenAI compatible API | | |
216+
| Variable | Description | Required? | Default |
217+
| ------------------------ | ------------------------------------------------------------- | -------------------------- | ------------------------ |
218+
| `OPENAI_API_KEY` | API key for OpenAI | Required when using OpenAI | |
219+
| `OPENAI_MODEL_NAME` | OpenAI model name to use | | `gpt-4.1` |
220+
| `OPENAI_EMBEDDING_MODEL` | OpenAI embedding model for verified answers | | `text-embedding-3-small` |
221+
| `OPENAI_BASE_URL` | Optional base URL for OpenAI compatible API | | |
222+
| `OPENAI_AVAILABLE_MODELS`| Comma-separated list of models available in the model picker | | All supported models |
216223

217224
#### Anthropic Configuration
218225

219-
| Variable | Description | Required? | Default |
220-
| ---------------------- | --------------------------- | ----------------------------- | -------------------------- |
221-
| `ANTHROPIC_API_KEY` | API key for Anthropic | Required when using Anthropic | |
222-
| `ANTHROPIC_MODEL_NAME` | Anthropic model name to use | | `claude-4-sonnet-20250514` |
226+
| Variable | Description | Required? | Default |
227+
| -------------------------- | ------------------------------------------------------------ | ----------------------------- | ------------------ |
228+
| `ANTHROPIC_API_KEY` | API key for Anthropic | Required when using Anthropic | |
229+
| `ANTHROPIC_MODEL_NAME` | Anthropic model name to use | | `claude-sonnet-4-5`|
230+
| `ANTHROPIC_AVAILABLE_MODELS`| Comma-separated list of models available in the model picker | | All supported models |
223231

224232
#### Azure AI Configuration
225233

@@ -239,6 +247,35 @@ The AI Analyst supports multiple providers for flexibility. Choose one of the pr
239247
| `OPENROUTER_SORT_ORDER` | Provider sorting method (`price`, `throughput`, `latency`) | | `latency` |
240248
| `OPENROUTER_ALLOWED_PROVIDERS` | Comma-separated list of allowed providers (`anthropic`, `openai`, `google`) | | `openai` |
241249

250+
#### AWS Bedrock Configuration
251+
252+
| Variable | Description | Required? | Default |
253+
| -------------------------- | ------------------------------------------------------------- | ------------------------------------------------ | --------------------- |
254+
| `BEDROCK_API_KEY` | API key for Bedrock (alternative to IAM credentials) | Required if not using IAM credentials | |
255+
| `BEDROCK_ACCESS_KEY_ID` | AWS access key ID for Bedrock | Required if not using API key | |
256+
| `BEDROCK_SECRET_ACCESS_KEY`| AWS secret access key for Bedrock | Required if using access key ID | |
257+
| `BEDROCK_SESSION_TOKEN` | AWS session token (for temporary credentials) | | |
258+
| `BEDROCK_REGION` | AWS region for Bedrock | Required when using IAM credentials | |
259+
| `BEDROCK_MODEL_NAME` | Bedrock model name to use | | `claude-sonnet-4-5` |
260+
| `BEDROCK_EMBEDDING_MODEL` | Bedrock embedding model for verified answers | | `cohere.embed-english-v3` |
261+
| `BEDROCK_AVAILABLE_MODELS` | Comma-separated list of models available in the model picker | | All supported models |
262+
263+
#### Supported Models
264+
265+
**OpenAI:** `gpt-5.1`, `gpt-4.1`
266+
267+
**Anthropic:** `claude-sonnet-4-5`, `claude-haiku-4-5`, `claude-sonnet-4`
268+
269+
**AWS Bedrock:** `claude-sonnet-4-5`, `claude-haiku-4-5`, `claude-sonnet-4`
270+
271+
<Note>
272+
Exact model snapshots are automatically assigned (e.g., `gpt-5.1` → `gpt-5.1-2025-11-13`).
273+
</Note>
274+
275+
<Note>
276+
For Bedrock, the region prefix is also added based on `BEDROCK_REGION` (e.g., `claude-sonnet-4-5` → `us.anthropic.claude-sonnet-4-5-20250929-v1:0`).
277+
</Note>
278+
242279
## Slack Integration
243280

244281
These variables enable you to configure the [Slack integration](/guides/using-slack-integration).

0 commit comments

Comments
 (0)