OpenMates Docs Open Chat

Daily Inspiration

Daily Inspiration Delivers a daily 10-card mixed inspiration set: 3 videos, 3 Wikipedia article prompts, and 4 OpenMates feature tips, with zero LLM latency ...

[T:documentation.sender_name]

Daily Inspiration

Delivers a daily 10-card mixed inspiration set: 3 videos, 3 Wikipedia article prompts, and 4 OpenMates feature tips, with zero LLM latency at display time.

Why This Exists

Provides curiosity-driven content from independent educators, journalists, documentary makers, and Wikipedia articles, plus static OpenMates feature tips. Content inspirations include a phrase, pre-written assistant response, and follow-up suggestions so users can start a chat instantly. Logged-out product feature cards use concise editorial headlines from product_features.yml; authenticated feature tips can deep-link to settings instead of creating chats.

How It Works

Generation Pipeline

Runs in app-ai-worker Celery container:

  1. Topic filtering – removes OpenMates refs, sensitive content, corporate/greenwashing framing via keyword blocklists in generator.py
  2. Brave Video Search – appends “educational independent creator” to bias toward independents; safesearch=strict, 20 results per slot
  3. YouTube-only + family-friendly filter – rejects non-YouTube and family_friendly=False
  4. Anti-corporate filter – seed blocklist (119 patterns in corporate_channel_patterns.yml) + Mistral Small LLM classifier for remaining channels (fails open)
  5. YouTube Data API enrichment – view count, duration, published_at, made-for-kids status
  6. Made-for-kids filter – rejects videos YouTube marks as Made for Kids before ranking or LLM selection
  7. Sort by view count – top 20 candidates per slot
  8. Main LLM call – single Mistral Small call for the 3 video slots; selects best videos and writes phrase/title/assistant_response/category/follow_up_suggestions
  9. Wiki generation – one cheap LLM call proposes Wikipedia article titles from the same recent topic pool, validates them via batch_validate_topics(), then fetches summaries through the Wikimedia provider
  10. Feature tips – four static OpenMates feature cards from feature_suggestions.py, no LLM/API cost
  11. Daily order – the 10-card set is shuffled once per user per UTC day before delivery and persisted in that order

Generation Triggers

Trigger When Count File
First paid request After first paid request post-processing 10 daily_inspiration_tasks.py
Scheduled daily 06:00 UTC Celery Beat 10 if viewed_count > 0 daily_inspiration_tasks.py
Default selection 06:30 UTC Celery Beat Mixed top 10 per language default_inspiration_tasks.py

Delivery

Authenticated users: Generated -> stored in Redis (7-day TTL) -> broadcast via WebSocket if online -> delivered on login if offline. Cross-device persistence via encrypted Directus user_daily_inspirations table.

Unauthenticated web users: The new-chat landing keeps its local six-slide OpenMates onboarding carousel. This guest-only source is never persisted as authenticated Daily Inspiration data.

Public and authenticated fallback clients: At 06:30 UTC, pool entries are scored per language (interaction_count / (age_hours + 1)) and a 3-video, 3-wiki, 4-feature target is written to daily_inspiration_defaults. GET /v1/default-inspirations?lang={code} exposes the credit-free public fallback with a 1-hour cache. Authenticated clients use it only when personalized IndexedDB, Phase 1, live WebSocket, and account API recovery provide no usable records.

Frontend

  • DailyInspirationBanner.svelte – carousel of up to 10 cards
  • dailyInspirationStore.ts – Svelte store for state/navigation and explicit guest/personalized/public/fallback source priority
  • dailyInspirationDB.ts – IndexedDB with AES-GCM encryption, 72h TTL
  • product_features.yml – logged-out OpenMates feature headline registry
  • Chat creation: handleStartChatFromInspiration() in ActiveChat.svelte

Data Structures

Pydantic Models (schemas.py)

  • DailyInspirationVideo – youtube_id, title, thumbnail_url, channel_name, view_count, duration_seconds, published_at
  • DailyInspirationWiki – title, wiki_title, description, thumbnail_url, wikidata_id, extract
  • DailyInspirationFeature – feature_id, icon, title, description, settings_path
  • DailyInspiration – inspiration_id, phrase, title, assistant_response, category, content_type, video, wiki, feature, generated_at, follow_up_suggestions

Directus Tables

Table Purpose
daily_inspiration_pool Shared cleartext pool, capped at 100
daily_inspiration_defaults Top 3 per language per day
user_daily_inspirations Per-user encrypted records

Cache Keys (Redis via cache_inspiration_mixin.py)

Key TTL Purpose
Topic suggestions 72h 3-day rolling window of user interests
Paid request tracking 48h Eligibility for next-day generation
Pending delivery 7 days Offline user inspiration queue
Sync cache 10 min Deduplication during multi-device sync

Edge Cases

  • LLM classifier failure: Fails open – candidates proceed unfiltered rather than dropping all content
  • Video cost remains capped: existing 3-video/day budget is unchanged
  • Wiki cost: one small LLM call plus free Wikimedia API requests
  • Feature cost: static/no external API calls
  • Message Processing – topic extraction via post-processing
  • E2E test: frontend/apps/web_app/tests/daily-inspiration-chat-flow.spec.ts