Media & Spirituality United States

Faithwave

Christian Media & Prayer Platform

Faithwave tablet layout in dark mode showing prayer library with expanded navigation
01 Overview

About the Project

Faithwave is an audio-first Flutter mobile app that pairs a curated Christian content library (prayers, abridged Christian classics, expert conversations, and multi-day prayer courses) with a custom audio engine that mixes a narrated track with optional looping ambient music. A conversational Prayer Finder uses semantic search over the full content catalog to recommend prayers and readings in natural language. Faithwave is rounded out by a web player that lets paying members share content with friends through time-limited leases, and a marketing site for acquisition.

Industry Media & Spirituality
Location United States
Team 2 Mobile Engineers, 1 Backend Engineer, 1 AI Engineer, 1 Audio Producer

The Challenge

Two thousand years of Christian devotional and theological writing is hard to navigate as audio: catalogs are huge, voice production is uneven, and traditional search ('I need a prayer for grief at 2am') matches keywords rather than spiritual intent. Faithwave needed a mobile experience that felt like a streaming app (multi-voice narration with optional ambient music, offline listening, multi-day course progress, sharing) backed by an editorial CMS and an AI assistant that could understand spiritual context instead of just matching words. It also had to span Christian denominations and be operable by a small team on a tight infra budget.

02 Solution

Our Approach

We built the mobile app in Flutter, with a Bloc/Cubit architecture and a custom playback engine on top of just_audio and audio_service that synchronises a narrated track with a looping ambient background. Auth is handled via Google and Apple Sign-In, paywall via RevenueCat, push via OneSignal and Firebase, and product analytics via PostHog. The backend is a Go service built on PocketBase, deployed alongside Ollama for embeddings (nomic-embed-text) and OpenAI's gpt-4o-mini for chat. The Prayer Finder is a RAG pipeline over a sqlite-vec vector index of all transcripts, rebuilt every 30 seconds by a background indexer, letting users describe a need in plain language and get back specific recommended prayers. Shared content is delivered through a separate Astro and React web player that streams audio over HLS and enforces time-limited content leases for trial users.

04 Technology

Tech Stack

The technologies and tools that powered this project.

Flutter
Dart
Bloc / Cubit
Go
PocketBase
SQLite + sqlite-vec
Ollama (nomic-embed-text)
OpenAI (gpt-4o-mini)
AWS S3
HLS / hls.js
Astro
React
Next.js
Tailwind CSS
RevenueCat
Firebase
OneSignal
PostHog
05 Results
500+ Prayers in Library
4.9★ App Store Rating
12 min Avg. Daily Session

Project Timeline

Content Strategy & Architecture

3 weeks

Content taxonomy design, audio pipeline architecture, AI prayer finder proof-of-concept with RAG, and UX research with faith community members.

Core App Development

8 weeks

Audio playback engine with multi-voice support, content browsing and discovery, user accounts, progress tracking, and offline download system.

AI Chat & Content Pipeline

4 weeks

RAG-based prayer finder implementation, vector database setup, content embedding pipeline, and editorial CMS for ongoing content publication.

Audio Production & Launch

4 weeks

Initial content catalog recording and production, beta testing with faith communities, App Store optimization, and launch preparation.

Team Composition

2 Mobile Engineers, 1 Backend Engineer, 1 AI Engineer, 1 Audio Producer

Let's build something great together

Ready to turn your idea into a product users love? Book a discovery call and meet the team that will build it.

Book a Call