Queue audio URLs and track usage limits
Connect
SPIRITT
DeepgramDeepgram
Audio into action
SPIRITT uses Deepgram to transcribe recorded audio and turn text into speech. Build searchable recording libraries, operate audio publishing workflows, and track project usage in one workspace.
What you can do with Deepgram.
Recordings become action
Convert accessible audio recordings into transcripts, then carry decisions, customer requests, and assigned work into your connected tools.
Written content gets heard
Generate speech from finished text and build listening experiences around training material, product guidance, and customer education.
Usage stays visible
Bring Deepgram project usage, permission scopes, and model metadata into practical dashboards and operational checks.
Works with your other apps.
Get started in three steps.
01
Open your workspace
Start a SPIRITT workspace for your recording library, spoken content, or audio operations.
02
Connect Deepgram
Connect Deepgram with an API key, identify the projects you want to use, and check their permission scopes.
03
Delegate an audio mission
Provide publicly accessible recording URLs or text to voice. SPIRITT builds the workflow and delivers the transcripts, audio, or connected system.
Hand over a mission.
Choose a mission. Make it yours.
Put Deepgram to work.
Give every recording a next step
Questions
Deepgram, with SPIRITT.
01Can SPIRITT transcribe live calls through this connection?+
The available transcription action processes pre-recorded audio from a publicly accessible URL. Use it for recordings you are authorized to process. Live audio streaming is not part of the supplied connection, and SPIRITT does not need to join a meeting to process its recording afterward.
02What access does Deepgram need?+
Deepgram connects with an API key. SPIRITT can list accessible projects and retrieve their permission scopes to establish the mission's access boundary. Recorded audio must also be reachable at a public URL; a private Drive link alone does not meet that input requirement.
03Can it turn our written content into audio?+
Yes. SPIRITT can send text to Deepgram's text-to-speech action and use the generated audio in a training portal, speech preview app, or audio help center. A surrounding workflow can retain source text versions and regenerate audio when the narration changes.
04How can transcripts drive work in other tools?+
SPIRITT can use the returned text to create research pages in Notion, carry explicit commitments into Asana, or group product requests into Linear issues. Source recording links and transcript excerpts keep those downstream actions traceable rather than turning an inferred summary into an unsupported fact.
05What can an ongoing Deepgram operations dashboard show?+
SPIRITT can retrieve project usage summaries, including total audio duration and billable duration, and retain snapshots for trend comparisons. It can combine these with your thresholds in Google Sheets and send Slack alerts. Duration measures should not be presented as invoice totals without separate pricing and billing data.
06Does listing think models create a voice agent?+
No. Think model discovery returns available model information for planning. Public model listings and model details also help build selection catalogs, but discovery alone does not configure or run a voice agent. The executable audio actions supplied here are recorded-audio transcription and text-to-speech generation.