Architecture

AI Processing Pipeline

The AI Processing Pipeline is a set of cloud‑based services that run behind the scenes to make sermons easier to discover and listen to.

What It Does

After a sermon is uploaded through the Admin Tool, the AI pipeline:

1. Transcribes Audio
   • Converts the sermon audio into text using speech‑to‑text technology.
2. Generates a Summary
   • Produces a concise, readable summary of the message.
3. Assigns Topical Tags
   • Attaches meaningful tags (e.g., “hope”, “forgiveness”, “marriage”) based on the content.
4. Builds Waveform Data
   • Prepares the data needed to draw the audio waveform in the mobile app player.
5. Keeps the Podcast Feed Updated
   • Adds or updates podcast episodes in a feed that podcast apps can subscribe to.

Diagram

   Admin uploads audio
   via Admin Tool
           |
           v
   +--------------+      +-----------+
   |  Admin Tool  | ---> | Global API|
   +--------------+      +-----+-----+
                               |
                        data   |   audio links
                               v
                         +-----------+
                         | DB & S3   |
                         +-----+-----+
                               |
                               v
                    +------------------------+
                    |  AI Processing         |
                    |  Pipeline              |
                    +-----------+------------+
                                |
        +-----------------------+------------------------------+
        v                       v                              v
  Transcript & tags     Enriched sermon data          Podcast RSS feed
                        (summary, tags, waveform)            |
                               |                            v
                               v                     Podcast platforms
                      Mobile App & Website
                    (read enriched data)

Where It Runs

• Implemented as serverless functions in the cloud.
• Triggered automatically when new or updated content is ready.
• Communicates with the Global API’s database to read and update sermon records.

Example End-to-End Flow

1. A staff member uploads a new sermon through the Admin Tool.
2. The Global API stores basic details and an audio URL.
3. An AWS Lambda (or similar serverless function) is triggered to:
   • Download the audio
   • Call a speech-to-text service to create a transcript
4. Another step in the pipeline uses the transcript to:
   • Generate a concise summary in plain language
   • Suggest topical tags based on the content
5. The pipeline writes the enriched data (summary, tags, waveform info) back to the Global API’s database.
6. A final step updates the podcast RSS feed, which podcast apps read on their own schedule.

From the church’s point of view, this all happens automatically after the single upload step.

Why It Matters

This pipeline:

• Saves staff time by automating summaries and tagging.
• Makes it easier for people to find sermons based on their needs (e.g., anxiety, marriage, finances).
• Powers a richer listening experience in the mobile app and podcast platforms.

Key Technologies (High-Level)

At a high level, the pipeline uses:

• AWS Lambda or equivalent serverless compute
• AWS S3 for storing and reading audio files
• A speech-to-text service (e.g., Whisper API) to create transcripts
• LLM-based summarization and tagging to enrich the content
• The Global API and its database as the single source of truth for enriched sermon data

The detailed implementation may evolve over time, but the core idea remains the same: automatically turn raw audio into rich, searchable content and keep all downstream experiences in sync.

From the outside, it’s invisible—but it’s a key part of what makes the Thrive ecosystem feel smart and cohesive.

This site is open source. Improve this page.