Repository navigation
Getting Started
Zhe Nie edited this page Oct 6, 2026
·
4 revisions
| Direct tool, language, or SDK | Use in DreamLog |
|---|---|
| Swift | iOS application language. |
| SwiftUI | Journal, recording, dream detail, and reflection screens. |
| Xcode | Build, run, debug, and test the iOS app. |
| AVFoundation | Record and play dream audio with native iOS APIs. |
Foundation URLSession |
Send saved audio and text to the back-end API. |
| SQLite3 | Keep local dream drafts and retry metadata on the device. |
| Direct tool, library, or API | Use in DreamLog |
|---|---|
| Python | Back-end language and processing jobs. |
| Starlette | HTTP endpoints for records, audio uploads, transcription jobs, and reflection threads. |
| Granian | ASGI server for the Starlette application. |
SQLite3 / Python sqlite3 |
Store dream records, revisions, job states, transcripts, summaries, and conversation turns. |
| OpenAI Python SDK | Call the OpenAI APIs below from the server; its automatically installed dependencies are omitted. |
| OpenAI Audio Transcriptions API | Convert recorded dream audio to editable text; gpt-4o-mini-transcribe is a candidate model. |
| OpenAI Responses API | Generate optional summaries, tags, and private reflection responses. |
| Add only if needed | Use |
|---|---|
| OpenAI Images API | Generate a visual for Arc D, if that stretch goal is adopted. |
| OpenAI Embeddings API | Turn dream text into vectors for Arc C semantic matching, if vector retrieval is chosen. |
| sqlite-vector | Search those vectors in SQLite3 for Arc C, if vector retrieval is chosen. |
The iOS app sends audio and reflection requests to the Python back end; only the server holds the OpenAI API key. Packages installed automatically as dependencies of the listed tools are intentionally omitted. See Project Architecture and Features for the proposed workflow and the remaining decisions.