Podcast Guest Intelligence API

Turn podcast RSS feeds into structured guest-appearance intelligence.

This Actor parses podcast feeds, extracts episode metadata, identifies likely guest names, companies, and roles, and scores each episode against your topic keywords. When the RSS text is weak, it can fall back to episode-page context to recover guest details. It is built for PR teams, founder-brand operators, recruiters, sales teams, and researchers who want a cleaner workflow than manually scanning podcast apps.

Best results come from interview-style feeds where guest names or roles appear in episode titles, descriptions, or episode pages.

What you get

• podcast name
• episode title
• episode URL
• publication timestamp
• likely guest name
• likely guest company
• likely guest role
• guest confidence score
• guest extraction method
• page-context enrichment flag
• guest count
• additional guests for multi-guest episodes
• matched topic keywords
• topic match count
• fit score
• stable record ID
• run summary

Best use cases

• founder and executive outreach
• PR opportunity tracking
• competitor media monitoring
• podcast prospecting for startup and SaaS teams
• research feeds for active voices in a niche

Input

{
  "feedUrls": [
    "https://rss.art19.com/masters-of-scale"
  ],
  "topicKeywords": ["ai", "startup", "founder", "scale", "leadership", "innovation", "growth", "business", "strategy", "venture", "investing", "finance", "technology", "sales", "product", "operations", "hiring", "culture", "robotics", "consumer"],
  "maxEpisodesPerFeed": 15,
  "maxConcurrency": 8,
  "timeoutSecs": 30,
  "emitOnlyMatchedTopics": false,
  "userAgent": "Soft But Savage research@example.com"
}

Output

Each dataset item is a structured guest-appearance record. The Actor is designed to help you find people, companies, and topics worth outreach or follow-up.

Each record now includes:

• record_id for stable dedupe and syncing
• normalized published_at timestamps
• guest_confidence and guest_extraction_method
• page_context_used when episode-page enrichment was attempted
• guest_count and additional_guests for episodes with multiple extracted guests
• topic_match_count
• episode_summary

The Actor also writes a RUN_SUMMARY record to the default key-value store with per-feed status and counts. That summary now includes total records, records with extracted guests, high-confidence guest records, and page-enriched records.

Reliability controls

• maxConcurrency limits simultaneous episode enrichment per feed. Lower it if podcast sites throttle requests.
• timeoutSecs controls feed and episode-page request timeouts.
• Invalid feed URLs fail early with an actionable error.
• If every feed fails, the Actor saves RUN_SUMMARY and exits as failed instead of silently succeeding with an empty dataset.
• If only some episodes fail, successful records are still pushed and the first episode errors are listed in RUN_SUMMARY.

Topic matching

topicKeywords controls what appears in topic_matches. The Actor now expands common variants for popular business, technology, finance, media, and operator topics. For example, scale matches scaled and scaling, founder matches founders, ai matches artificial intelligence, robotics matches robotaxi or autonomous, hiring matches talent, and consumer matches customer or users.

If you pass a very small custom keyword list, fewer topics will appear because the Actor only reports matches for the keywords you requested.

Common errors

• Invalid feed URL: the feed must start with http:// or https://.
• No podcast records were saved because every feed failed: check that feedUrls are public RSS URLs and that your userAgent is accepted by the source.
• Zero records with a successful run: usually means emitOnlyMatchedTopics was enabled and no episode text matched your topicKeywords.