We’re an AppFoundry partner building a vertical-specific AI routing engine. To improve our base models, we want to aggregate the intent recognition data from our clients’ Genesys Cloud orgs.
We love the native ‘Intent Miner’ tool, but we can’t figure out how to programmatically export the mined utterances and intent mappings via the API. The GET /api/v2/languageunderstanding/miner endpoints let us trigger a mining job, but the actual dataset generated by the miner seems locked in the UI. Is there a hidden API endpoint or a workaround to extract the raw CSV of mined intents so we can feed it into our centralized ML pipeline across multiple tenants?
Building custom desktops, we hit API walls constantly.
There is no public API to extract the results of an Intent Miner job directly. The platform is intentionally designed to keep those utterances isolated within the org for data privacy (GDPR/CCPA) reasons. Genesys doesn’t want AppFoundry apps siphoning off raw customer transcripts en masse. If you look at the network tab in the browser when you click ‘Export’ in the Intent Miner UI, it uses a private, undocumented v2.internal endpoint that will reject OAuth client credentials.
We had to navigate strict APAC data residency laws with this.
As As noted above, you can’t export the Miner results. However, if you are an AppFoundry partner, you shouldn’t be relying on Intent Miner for your raw data feed anyway. You should use the ‘Analytics Conversation Detail’ API to pull the raw transcripts from your clients’ orgs (assuming they grant you that OAuth scope), and then run your own intent mining algorithm on your side. Let Genesys Cloud handle the routing, but do the heavy ML lifting and data aggregation in your own secure tenant where you control the data pipeline.