Release 1.2 (#242)

* chore: improve podcast transcripts

* fix: remove date from insight - fixes #241

* fix: improve scrolling on source and insights - fixes #237

* chore: update esperanto to fix: #234

* chore: update esperanto to fix #226

* fix: process vectorization as subcommands to handle larger documents more gracefully - fix: #229

* feat: enable background job retry capabilities

* feat: reenable content types that were disabled during alpha version

* fix: remove unnecessary model caching causing many issues.

* feat: support multiple azure endpoints and keys just like openai compatible. Fixes #215

* docs: update azure variables

* chore: bump and update dependencies

.env.example

@@ -143,10 +143,28 @@ API_URL=http://localhost:5055
 # OPENAI_COMPATIBLE_API_KEY_TTS=
 
 # AZURE OPENAI
+# Generic configuration (applies to all modalities: language, embedding, STT, TTS)
 # AZURE_OPENAI_API_KEY=
 # AZURE_OPENAI_ENDPOINT=
-# AZURE_OPENAI_API_VERSION="2024-12-01-preview"
-# AZURE_OPENAI_DEPLOYMENT_NAME=
+# AZURE_OPENAI_API_VERSION=2024-12-01-preview
+
+# Mode-specific configuration (overrides generic if set)
+# Use these when you want different deployments for different AI capabilities
+# AZURE_OPENAI_API_KEY_LLM=
+# AZURE_OPENAI_ENDPOINT_LLM=
+# AZURE_OPENAI_API_VERSION_LLM=
+
+# AZURE_OPENAI_API_KEY_EMBEDDING=
+# AZURE_OPENAI_ENDPOINT_EMBEDDING=
+# AZURE_OPENAI_API_VERSION_EMBEDDING=
+
+# AZURE_OPENAI_API_KEY_STT=
+# AZURE_OPENAI_ENDPOINT_STT=
+# AZURE_OPENAI_API_VERSION_STT=
+
+# AZURE_OPENAI_API_KEY_TTS=
+# AZURE_OPENAI_ENDPOINT_TTS=
+# AZURE_OPENAI_API_VERSION_TTS=
 
 # USE THIS IF YOU WANT TO DEBUG THE APP ON LANGSMITH
 # LANGCHAIN_TRACING_V2=true
@@ -162,6 +180,63 @@ SURREAL_PASSWORD="root"
 SURREAL_NAMESPACE="open_notebook"
 SURREAL_DATABASE="staging"
 
+# RETRY CONFIGURATION (surreal-commands v1.2.0+)
+# Global defaults for all background commands unless explicitly overridden at command level
+# These settings help commands automatically recover from transient failures like:
+# - Database transaction conflicts during concurrent operations
+# - Network timeouts when calling external APIs
+# - Rate limits from LLM/embedding providers
+# - Temporary resource unavailability
+
+# Enable/disable retry globally (default: true)
+# Set to false to disable retries for all commands (useful for debugging)
+SURREAL_COMMANDS_RETRY_ENABLED=true
+
+# Maximum retry attempts before giving up (default: 3)
+# Database operations use 5 attempts (defined per-command)
+# API calls use 3 attempts (defined per-command)
+SURREAL_COMMANDS_RETRY_MAX_ATTEMPTS=3
+
+# Wait strategy between retry attempts (default: exponential_jitter)
+# Options: exponential_jitter, exponential, fixed, random
+# - exponential_jitter: Recommended - prevents thundering herd during DB conflicts
+# - exponential: Good for API rate limits (predictable backoff)
+# - fixed: Use for quick recovery scenarios
+# - random: Use when you want unpredictable retry timing
+SURREAL_COMMANDS_RETRY_WAIT_STRATEGY=exponential_jitter
+
+# Minimum wait time between retries in seconds (default: 1)
+# Database conflicts: 1 second (fast retry for transient issues)
+# API rate limits: 5 seconds (wait for quota reset)
+SURREAL_COMMANDS_RETRY_WAIT_MIN=1
+
+# Maximum wait time between retries in seconds (default: 30)
+# Database conflicts: 30 seconds maximum
+# API rate limits: 120 seconds maximum (defined per-command)
+# Total retry time won't exceed max_attempts * wait_max
+SURREAL_COMMANDS_RETRY_WAIT_MAX=30
+
+# WORKER CONCURRENCY
+# Maximum number of concurrent tasks in the worker pool (default: 5)
+# This affects the likelihood of database transaction conflicts during batch operations
+#
+# Tuning guidelines based on deployment size:
+# - Resource-constrained (low CPU/memory): 1-2 workers
+#   Reduces conflicts and resource usage, but slower processing
+#
+# - Normal deployment (balanced): 5 workers (RECOMMENDED)
+#   Good balance between throughput and conflict rate
+#   Retry logic handles occasional conflicts gracefully
+#
+# - Large instances (high CPU/memory): 10-20 workers
+#   Higher throughput but more frequent DB conflicts
+#   Relies heavily on retry logic with jittered backoff
+#
+# Note: Higher concurrency increases vectorization speed but also increases
+# SurrealDB transaction conflicts. The retry logic with exponential-jitter
+# backoff ensures operations complete successfully even at high concurrency.
+SURREAL_COMMANDS_MAX_TASKS=5
+
 # OPEN_NOTEBOOK_PASSWORD=
 
 # FIRECRAWL - Get a key at https://firecrawl.dev/