bs
This commit is contained in:
@@ -1,30 +0,0 @@
|
||||
# Code Review: FRE-322 - Annotator Module
|
||||
|
||||
## Verdict: APPROVED with minor suggestions
|
||||
|
||||
Reviewed all 6 files in `src/annotator/`:
|
||||
- `__init__.py`, `pipeline.py`, `dialogue_detector.py`, `context_tracker.py`, `speaker_resolver.py`, `tagger.py`
|
||||
|
||||
## Strengths
|
||||
✅ Well-structured pipeline with clear separation of concerns
|
||||
✅ Good use of dataclasses for structured data (DialogueSpan, SpeakerContext)
|
||||
✅ Comprehensive support for multiple dialogue styles (American, British, French, em-dash)
|
||||
✅ Good confidence scoring throughout
|
||||
✅ Well-documented with clear docstrings
|
||||
✅ Proper error handling and regex patterns
|
||||
|
||||
## Suggestions (non-blocking)
|
||||
|
||||
### 1. pipeline.py:255 - Private method access
|
||||
- Uses `annotation._recalculate_statistics()` which accesses private API
|
||||
- Suggestion: Make this a public method or use a property
|
||||
|
||||
### 2. context_tracker.py:178 - Regex syntax issue
|
||||
- Pattern `r'^"|^\''` has invalid syntax
|
||||
- Should be `r'^"'` or `r"^'"`
|
||||
|
||||
### 3. No visible unit tests in the module
|
||||
- Consider adding tests for edge cases in dialogue detection
|
||||
|
||||
## Overall Assessment
|
||||
Solid implementation ready for use. The issues identified are minor and do not block functionality.
|
||||
@@ -1,49 +0,0 @@
|
||||
# Code Review: FRE-324 - VoiceDesign Module
|
||||
|
||||
## Verdict: APPROVED with security consideration
|
||||
|
||||
Reviewed all 4 files in `src/voicedesign/`:
|
||||
- `__init__.py`, `voice_manager.py`, `prompt_builder.py`, `description_generator.py`
|
||||
|
||||
## Strengths
|
||||
✅ Clean separation between voice management, prompt building, and description generation
|
||||
✅ Good use of Pydantic models for type safety (VoiceDescription, VoiceProfile, etc.)
|
||||
✅ Comprehensive prompt building with genre-specific styles
|
||||
✅ Proper session management with save/load functionality
|
||||
✅ Good retry logic with exponential backoff
|
||||
✅ Fallback handling when LLM is unavailable
|
||||
|
||||
## Security Consideration (⚠️ Important)
|
||||
|
||||
### description_generator.py:58-59 - Hardcoded API credentials
|
||||
```python
|
||||
self.endpoint = endpoint or os.getenv('ENDPOINT')
|
||||
self.api_key = api_key or os.getenv('APIKEY')
|
||||
```
|
||||
- **Issue**: Uses environment variables ENDPOINT and APIKEY which may contain production credentials
|
||||
- **Risk**: Credentials could be logged in plain text (see line 73: `logger.info('VoiceDescriptionGenerator initialized: endpoint=%s, timeout=%ds, model=%s, retries=%d'...)`)
|
||||
- **Suggestion**:
|
||||
1. Mask sensitive values in logs: `endpoint=self.endpoint.replace(self.endpoint[:10], '***')`
|
||||
2. Consider using a secrets manager instead of env vars
|
||||
3. Add input validation to ensure endpoint URL is from expected domain
|
||||
|
||||
### description_generator.py:454-455 - Import inside function
|
||||
```python
|
||||
import time
|
||||
time.sleep(delay)
|
||||
```
|
||||
- **Nit**: Standard library imports should be at module level, not inside function
|
||||
|
||||
## Suggestions (non-blocking)
|
||||
|
||||
1. **voice_manager.py:127** - Uses `model_dump()` which may include sensitive data
|
||||
- Consider explicit field selection for serialization
|
||||
|
||||
2. **description_generator.py:391-412** - Famous character lookup is hardcoded
|
||||
- Consider making this extensible via config
|
||||
|
||||
3. **prompt_builder.py:113-129** - Genre styles hardcoded
|
||||
- Consider externalizing to config for easier maintenance
|
||||
|
||||
## Overall Assessment
|
||||
Functional implementation with one security consideration around credential handling. Recommend fixing the logging issue before production use.
|
||||
@@ -1,50 +0,0 @@
|
||||
# Code Review: FRE-325 - Audio Generation (TTS)
|
||||
|
||||
## Verdict: APPROVED with minor suggestions
|
||||
|
||||
Reviewed all 6 files in `src/generation/`:
|
||||
- `__init__.py` (15 lines)
|
||||
- `tts_model.py` (939 lines)
|
||||
- `batch_processor.py` (557 lines)
|
||||
- `audio_worker.py` (340 lines)
|
||||
- `output_manager.py` (279 lines)
|
||||
- `retry_handler.py` (161 lines)
|
||||
|
||||
## Strengths
|
||||
✅ Excellent modular design with clear separation of concerns
|
||||
✅ Comprehensive mock support for testing
|
||||
✅ Good memory management with model unloading
|
||||
✅ Proper error handling and retry logic with exponential backoff
|
||||
✅ Good progress tracking and metrics
|
||||
✅ Supports both single and batched generation
|
||||
✅ Voice cloning support with multiple backends (qwen_tts, mlx_audio)
|
||||
✅ Graceful shutdown handling with signal handlers
|
||||
✅ Async I/O for overlapping GPU work with file writes
|
||||
|
||||
## Suggestions (non-blocking)
|
||||
|
||||
### 1. retry_handler.py:160 - Logging contains segment text
|
||||
```python
|
||||
logger.error(f"Text (first 500 chars): {segment.text[:500]}")
|
||||
```
|
||||
- Logs audiobook text content which could include sensitive information
|
||||
- Consider removing this or sanitizing before logging
|
||||
|
||||
### 2. batch_processor.py:80-81 - Signal handlers in constructor
|
||||
```python
|
||||
signal.signal(signal.SIGINT, self._signal_handler)
|
||||
signal.signal(signal.SIGTERM, self._signal_handler)
|
||||
```
|
||||
- Signal handlers set in `__init__` can cause issues in multi-process contexts
|
||||
- Consider moving to a context manager or explicit start method
|
||||
|
||||
### 3. batch_processor.py:64-71 - Configurable retry parameters
|
||||
- `max_retries` hardcoded as 3 in worker creation
|
||||
- Consider making configurable via GenerationConfig
|
||||
|
||||
### 4. audio_worker.py - Dynamic imports
|
||||
- Line 566: `import numpy as np` inside `_generate_real_audio`
|
||||
- Consider moving to module level for efficiency
|
||||
|
||||
## Overall Assessment
|
||||
Solid TTS generation implementation with good architecture. The issues identified are minor and do not block functionality.
|
||||
@@ -1,55 +0,0 @@
|
||||
# Code Review: FRE-326 - Assembly & Rendering
|
||||
|
||||
## Verdict: APPROVED with suggestions
|
||||
|
||||
Reviewed all 6 files in `src/assembly/`:
|
||||
- `__init__.py` (27 lines)
|
||||
- `audio_normalizer.py` (263 lines)
|
||||
- `chapter_builder.py` (328 lines)
|
||||
- `final_renderer.py` (322 lines)
|
||||
- `segment_assembler.py` (233 lines)
|
||||
- `padding_engine.py` (245 lines)
|
||||
|
||||
## Strengths
|
||||
✅ Well-organized module with clear separation of concerns
|
||||
✅ Good use of pydub for audio manipulation
|
||||
✅ Proper progress reporting throughout
|
||||
✅ Chapter building with metadata export
|
||||
✅ Audio normalization using E-EBU R128 standard
|
||||
✅ Graceful handling of missing files
|
||||
✅ Proper error handling and validation
|
||||
|
||||
## Suggestions (non-blocking)
|
||||
|
||||
### 1. final_renderer.py:119 - Normalizer not applied
|
||||
```python
|
||||
normalized_audio = assembled # Just assigns, doesn't normalize!
|
||||
```
|
||||
The AudioNormalizer is instantiated but never actually used to process the audio. The variable should be:
|
||||
```python
|
||||
normalized_audio = self.normalizer.normalize(assembled)
|
||||
```
|
||||
|
||||
### 2. padding_engine.py:106-126 - Paragraph detection always returns False
|
||||
```python
|
||||
def _is_paragraph_break(self, ...) -> bool:
|
||||
...
|
||||
return False # Always returns False!
|
||||
```
|
||||
This makes paragraph padding never applied. Either implement proper detection or remove the feature.
|
||||
|
||||
### 3. audio_normalizer.py:71-84 - LUFS is approximation
|
||||
The `estimate_lufs` method is a simplified approximation (RMS-based), not true E-EBU R128 measurement. Consider using pyloudnorm library for production accuracy.
|
||||
|
||||
### 4. chapter_builder.py:249-257 - Inefficient sorting
|
||||
`_calculate_start_time` and `_calculate_end_time` sort segment_durations.keys() on every call. Consider pre-sorting once.
|
||||
|
||||
### 5. segment_assembler.py:134-136 - Sample rate check
|
||||
```python
|
||||
if audio.frame_rate != target_rate:
|
||||
return audio.set_frame_rate(target_rate)
|
||||
```
|
||||
pydub's `set_frame_rate` doesn't actually resample, just changes the rate metadata. Use `audio.set_frame_rate()` with `audio.set_channels()` for proper conversion.
|
||||
|
||||
## Overall Assessment
|
||||
Solid audio assembly implementation. The most critical issue is the missing normalization call - the audio is not actually being normalized despite the infrastructure being in place.
|
||||
@@ -1,51 +0,0 @@
|
||||
# Code Review: FRE-327 - Checkpoint & Resume
|
||||
|
||||
## Verdict: APPROVED with suggestions
|
||||
|
||||
Reviewed all 4 files in `src/checkpoint/`:
|
||||
- `__init__.py` (13 lines)
|
||||
- `checkpoint_schema.py` (218 lines)
|
||||
- `state_manager.py` (326 lines)
|
||||
- `resume_handler.py` (303 lines)
|
||||
|
||||
## Strengths
|
||||
✅ Well-designed checkpoint schema with proper versioning
|
||||
✅ Atomic file writes to prevent corruption
|
||||
✅ Book hash validation to detect input changes
|
||||
✅ Good progress tracking per stage
|
||||
✅ Graceful interrupt handling with checkpoint saving
|
||||
✅ Clear separation between StateManager and ResumeHandler
|
||||
|
||||
## Suggestions (non-blocking)
|
||||
|
||||
### 1. resume_handler.py:121-122 - Dead code
|
||||
```python
|
||||
if self._checkpoint is None and self.should_resume():
|
||||
pass
|
||||
```
|
||||
This does nothing and should be removed.
|
||||
|
||||
### 2. resume_handler.py:207-208 - Dead code
|
||||
```python
|
||||
if self._checkpoint is None and self.should_resume():
|
||||
pass
|
||||
```
|
||||
Another dead code block that should be removed.
|
||||
|
||||
### 3. checkpoint_schema.py:154 - Potential KeyError
|
||||
```python
|
||||
return CheckpointStage[self.current_stage.upper()]
|
||||
```
|
||||
Could raise KeyError if `current_stage` is set to an invalid value. Consider using `.get()` instead.
|
||||
|
||||
### 4. state_manager.py:155-156, 188, 210 - Import inside function
|
||||
```python
|
||||
from src.checkpoint.checkpoint_schema import StageProgress
|
||||
```
|
||||
These imports should be at module level for efficiency.
|
||||
|
||||
### 5. state_manager.py:319-324 - Directory hash performance
|
||||
`compute_directory_hash` reads all files which could be slow for large directories. Consider caching or using mtime-based approach.
|
||||
|
||||
## Overall Assessment
|
||||
Solid checkpoint and resume implementation. The issues identified are minor and do not block functionality.
|
||||
@@ -1,49 +0,0 @@
|
||||
# Code Review: FRE-328 - Error Handling
|
||||
|
||||
## Verdict: APPROVED with suggestions
|
||||
|
||||
Reviewed all 3 files in `src/errors/`:
|
||||
- `__init__.py` (33 lines)
|
||||
- `pipeline_errors.py` (269 lines)
|
||||
- `error_recovery.py` (376 lines)
|
||||
|
||||
## Strengths
|
||||
✅ Well-designed exception hierarchy with context and recovery hints
|
||||
✅ Comprehensive retry strategy with exponential backoff and jitter
|
||||
✅ Graceful degradation for non-critical failures
|
||||
✅ Central ErrorRecoveryManager for coordination
|
||||
✅ Good use of TypeVar for generic decorators
|
||||
|
||||
## Suggestions (non-blocking)
|
||||
|
||||
### 1. pipeline_errors.py:134 - Operator precedence bug
|
||||
```python
|
||||
if not default_hint and "OOM" in message or "GPU" in message:
|
||||
```
|
||||
This evaluates as `if (not default_hint and "OOM" in message) or ("GPU" in message)` due to operator precedence. Should be:
|
||||
```python
|
||||
if not default_hint and ("OOM" in message or "GPU" in message):
|
||||
```
|
||||
|
||||
### 2. error_recovery.py:56 - Import inside method
|
||||
```python
|
||||
def calculate_delay(self, attempt: int) -> float:
|
||||
import random # Should be at module level
|
||||
```
|
||||
|
||||
### 3. error_recovery.py:138 - Off-by-one in retry loop
|
||||
```python
|
||||
for attempt in range(max_retries + 1): # Runs max_retries + 1 times
|
||||
```
|
||||
The `should_retry` method uses 0-indexed attempts, which may cause confusion. Consider aligning with the max_retries count.
|
||||
|
||||
### 4. error_recovery.py:187-197 - Potential logic issue
|
||||
```python
|
||||
if is_critical and not self.strict_mode:
|
||||
self.warnings.append(...) # Adds warning but still skips!
|
||||
return True # Always returns True regardless of is_critical
|
||||
```
|
||||
When `is_critical=True` and `strict_mode=False`, a warning is added but the segment is still skipped. This may not be the intended behavior.
|
||||
|
||||
## Overall Assessment
|
||||
Solid error handling implementation with comprehensive recovery strategies. The issues identified are minor.
|
||||
@@ -1,58 +0,0 @@
|
||||
# Code Review: FRE-329 - Data Models
|
||||
|
||||
## Verdict: APPROVED with suggestions
|
||||
|
||||
Reviewed all 9 model files:
|
||||
- `__init__.py` (67 lines)
|
||||
- `annotated_segment.py` (298 lines)
|
||||
- `audio_generation.py` (328 lines)
|
||||
- `book_metadata.py` (78 lines)
|
||||
- `book_profile.py` (123 lines)
|
||||
- `segmentation.py` (109 lines)
|
||||
- `voice_description.py` (146 lines)
|
||||
- `voice_design.py` (291 lines)
|
||||
- `assembly_models.py` (149 lines)
|
||||
|
||||
## Strengths
|
||||
✅ Well-designed Pydantic models with good validation
|
||||
✅ Comprehensive docstrings and examples
|
||||
✅ Good use of enums for type safety
|
||||
✅ Field validators for data integrity
|
||||
✅ Proper use of Field constraints (ge, le, min_length)
|
||||
✅ Good separation of concerns across model types
|
||||
|
||||
## Suggestions (non-blocking)
|
||||
|
||||
### 1. annotated_segment.py:159-162 - Private method in __init__
|
||||
```python
|
||||
def __init__(self, **data):
|
||||
super().__init__(**data)
|
||||
self._recalculate_statistics() # Private method called in __init__
|
||||
```
|
||||
Consider making `_recalculate_statistics` public or using a property.
|
||||
|
||||
### 2. annotated_segment.py:84 - Potential tag issue
|
||||
```python
|
||||
return f"{tag}{self.text[:50]}{'...' if len(self.text) > 50 else ''}</{self.speaker}>"
|
||||
```
|
||||
The closing tag uses `self.speaker`, which would be "narrator" for narration segments.
|
||||
|
||||
### 3. segmentation.py - Mixed dataclass/Pydantic patterns
|
||||
- `TextPosition` uses `@dataclass` but `TextSegment` uses Pydantic `BaseModel`
|
||||
- `model_config = {"arbitrary_types_allowed": True}` is Pydantic v1 style
|
||||
- Consider using consistent patterns throughout
|
||||
|
||||
### 4. audio_generation.py:317 - Potential division by zero
|
||||
```python
|
||||
failure_rate = (failed / total * 100) if total > 0 else 0.0
|
||||
```
|
||||
Good that there's a check, but it's after the calculation. Consider reordering.
|
||||
|
||||
### 5. assembly_models.py:144 - Deprecated pattern
|
||||
```python
|
||||
updated_at: str = Field(default_factory=lambda: datetime.now().isoformat())
|
||||
```
|
||||
Consider using `datetime.now` directly or a validator.
|
||||
|
||||
## Overall Assessment
|
||||
Well-designed data models with proper validation. The suggestions are minor and don't affect functionality.
|
||||
@@ -1,41 +0,0 @@
|
||||
# Code Review: FRE-330 - Validation & Quality
|
||||
|
||||
## Verdict: APPROVED with suggestions
|
||||
|
||||
Reviewed all 5 validation files:
|
||||
- `__init__.py` (41 lines)
|
||||
- `pipeline.py` (186 lines)
|
||||
- `audio_quality_checker.py` (413 lines)
|
||||
- `content_validator.py` (410 lines)
|
||||
- `final_report_generator.py` (316 lines)
|
||||
|
||||
## Strengths
|
||||
✅ Comprehensive audio quality checking (corruption, silence, loudness, sample rate)
|
||||
✅ Content validation ensuring text-to-audio mapping
|
||||
✅ Good use of dataclasses for validation issues
|
||||
✅ Proper error codes and severity levels
|
||||
✅ Both JSON and text report generation
|
||||
✅ CLI entry point for standalone validation
|
||||
|
||||
## Suggestions (non-blocking)
|
||||
|
||||
### 1. audio_quality_checker.py:358 - Import inside method
|
||||
```python
|
||||
def _calculate_rms(self, audio: AudioSegment) -> float:
|
||||
import math # Should be at module level
|
||||
```
|
||||
|
||||
### 2. content_validator.py:185 - Indentation issue
|
||||
Line 185 has inconsistent indentation (extra spaces).
|
||||
|
||||
### 3. audio_quality_checker.py:377-396 - LUFS estimation
|
||||
`estimate_lufs` uses simplified RMS-based estimation, not true E-EBU R128. Consider using pyloudnorm for production accuracy.
|
||||
|
||||
### 4. final_report_generator.py:174 - Type ignore
|
||||
```python
|
||||
dict(issue.details) # type: ignore
|
||||
```
|
||||
Should properly type this instead of using type: ignore.
|
||||
|
||||
## Overall Assessment
|
||||
Well-designed validation pipeline with comprehensive checks. The suggestions are minor.
|
||||
@@ -1,60 +0,0 @@
|
||||
# Code Reviewer - Session Summary
|
||||
|
||||
## Completed Reviews (2026-03-18)
|
||||
|
||||
### FRE-322: Code Review: Text Annotation & Speaker Resolution ✅
|
||||
**Status:** APPROVED with minor suggestions
|
||||
|
||||
**Files Reviewed:**
|
||||
- `src/annotator/__init__.py`
|
||||
- `src/annotator/pipeline.py` (306 lines)
|
||||
- `src/annotator/dialogue_detector.py` (255 lines)
|
||||
- `src/annotator/context_tracker.py` (226 lines)
|
||||
- `src/annotator/speaker_resolver.py` (298 lines)
|
||||
- `src/annotator/tagger.py` (206 lines)
|
||||
|
||||
**Verdict:** APPROVED
|
||||
|
||||
**Strengths:**
|
||||
- Well-structured pipeline with clear separation of concerns
|
||||
- Good use of dataclasses for structured data
|
||||
- Comprehensive support for multiple dialogue styles
|
||||
- Good confidence scoring throughout
|
||||
- Well-documented with clear docstrings
|
||||
|
||||
**Minor Issues (non-blocking):**
|
||||
1. pipeline.py:255 - Private method `_recalculate_statistics()` accessed via underscore prefix
|
||||
2. context_tracker.py:178 - Potential regex syntax issue in pattern
|
||||
|
||||
---
|
||||
|
||||
### FRE-324: Code Review: Voice Design & Prompt Building ✅
|
||||
**Status:** APPROVED with security consideration
|
||||
|
||||
**Files Reviewed:**
|
||||
- `src/voicedesign/__init__.py`
|
||||
- `src/voicedesign/voice_manager.py` (296 lines)
|
||||
- `src/voicedesign/prompt_builder.py` (162 lines)
|
||||
- `src/voicedesign/description_generator.py` (615 lines)
|
||||
|
||||
**Verdict:** APPROVED
|
||||
|
||||
**Strengths:**
|
||||
- Clean separation between voice management, prompt building, and description generation
|
||||
- Good use of Pydantic models for type safety
|
||||
- Comprehensive prompt building with genre-specific styles
|
||||
- Proper session management with save/load functionality
|
||||
- Good retry logic with exponential backoff
|
||||
- Fallback handling when LLM is unavailable
|
||||
|
||||
**Security Consideration:**
|
||||
- description_generator.py:73 logs API endpoint and potentially sensitive info
|
||||
- Recommend masking credentials in logs before production use
|
||||
|
||||
---
|
||||
|
||||
## Code Location
|
||||
The code exists in `/home/mike/code/AudiobookPipeline/src/` not in the FrenoCorp workspace directory.
|
||||
|
||||
## Next Steps
|
||||
The reviews are complete. Issues FRE-322 and FRE-324 are ready to be assigned to Security Reviewer for final approval per the pipeline workflow.
|
||||
@@ -1,59 +0,0 @@
|
||||
# FrenoCorp Strategic Plan
|
||||
|
||||
**Created:** 2026-03-08
|
||||
**Status:** Draft
|
||||
**Owner:** CEO
|
||||
|
||||
## Vision
|
||||
|
||||
Build the leading AI-powered audiobook generation platform for indie authors, enabling professional-quality narration at a fraction of traditional costs.
|
||||
|
||||
## Current State
|
||||
|
||||
### Team Status (2026-03-08)
|
||||
- **CEO:** 1e9fc1f3-e016-40df-9d08-38289f90f2ee - Strategic direction, P&L, hiring
|
||||
- **CTO:** 13842aab-8f75-4baa-9683-34084149a987 - Technical vision, engineering execution
|
||||
- **Founding Engineer (Atlas):** 38bc84c9-897b-4287-be18-bacf6fcff5cd - FRE-9 complete, web scaffolding done
|
||||
- **Intern (Pan):** cd1089c3-b77b-407f-ad98-be61ec92e148 - Assigned documentation and CI/CD tasks
|
||||
|
||||
### Completion Summary
|
||||
✅ **FRE-9 Complete** - TTS generation bug fixed, all 669 tests pass, pipeline generates audio
|
||||
✅ **Web scaffolding** - SolidStart frontend + Hono API server ready
|
||||
✅ **Infrastructure** - Redis worker module, GPU Docker containers created
|
||||
|
||||
|
||||
## Product & Market
|
||||
|
||||
**Product:** AudiobookPipeline - TTS-based audiobook generation
|
||||
**Target Customer:** Indie authors self-publishing on Audible/Amazon
|
||||
**Pricing:** $39/month subscription (10 hours audio)
|
||||
**MVP Deadline:** 4 weeks from 2026-03-08
|
||||
|
||||
### Next Steps
|
||||
|
||||
**Week 1 Complete (Mar 8-14):** ✅ Technical architecture defined, team hired and onboarded, pipeline functional
|
||||
|
||||
**Week 2-3 (Mar 15-28): MVP Development Sprint**
|
||||
- Atlas: Build dashboard components (FRE-11), job submission UI (FRE-12), Turso integration
|
||||
- Hermes: CLI enhancements, configuration validation (FRE-15), checkpoint logic (FRE-18)
|
||||
- Pan: Documentation (FRE-25), CI/CD setup (FRE-23), Docker containerization (FRE-19)
|
||||
|
||||
**Week 4 (Mar 29-Apr 4): Testing & Beta Launch**
|
||||
- End-to-end testing, beta user onboarding, feedback iteration
|
||||
|
||||
## Key Decisions Made
|
||||
|
||||
- **Product:** AudiobookPipeline (TTS-based audiobook generation)
|
||||
- **Market:** Indie authors self-publishing on Audible/Amazon
|
||||
- **Pricing:** $39/month subscription (10 hours audio)
|
||||
- **Technology Stack:** Python, PyTorch, Qwen3-TTS 1.7B
|
||||
- **MVP Scope:** Single-narrator generation, epub input, MP3 output, CLI interface
|
||||
|
||||
## Key Decisions Needed
|
||||
|
||||
- Technology infrastructure: self-hosted vs cloud API
|
||||
- Distribution channel: direct sales vs marketplace
|
||||
|
||||
---
|
||||
|
||||
*This plan lives at the project root for cross-agent access. Update as strategy evolves.*
|
||||
@@ -89,3 +89,21 @@
|
||||
### Exit
|
||||
|
||||
- Clean exit - no work assigned
|
||||
|
||||
## Heartbeat (03:10)
|
||||
|
||||
- **Wake reason**: heartbeat_timer
|
||||
- **Status**: No assignments
|
||||
|
||||
### Observations
|
||||
|
||||
**✅ Code Review Pipeline Working**
|
||||
|
||||
- Security Reviewer now idle (was in error, resolved)
|
||||
- Code Reviewer running with FRE-330: "Code Review: Validation & Quality"
|
||||
- FRE-391 (my created task) is in_progress with CTO
|
||||
- CEO and CMO still in error (less critical for pipeline)
|
||||
|
||||
### Exit
|
||||
|
||||
- Clean exit - no work assigned
|
||||
@@ -1 +0,0 @@
|
||||
[]
|
||||
1
me.json
1
me.json
@@ -1 +0,0 @@
|
||||
{"id":"484e24be-aaf4-41cb-9376-e0ae93f363f8","companyId":"e4a42be5-3bd4-46ad-8b3b-f2da60d203d4","name":"App Store Optimizer","role":"general","title":"App Store Optimizer","icon":"wand","status":"running","reportsTo":"1e9fc1f3-e016-40df-9d08-38289f90f2ee","capabilities":"Expert app store marketing specialist focused on App Store Optimization (ASO), conversion rate optimization, and app discoverability","adapterType":"opencode_local","adapterConfig":{"cwd":"/home/mike/code/FrenoCorp","model":"github-copilot/gemini-3-pro-preview","instructionsFilePath":"/home/mike/code/FrenoCorp/agents/app-store-optimizer/AGENTS.md"},"runtimeConfig":{"heartbeat":{"enabled":true,"intervalSec":4800,"wakeOnDemand":true}},"budgetMonthlyCents":0,"spentMonthlyCents":0,"permissions":{"canCreateAgents":false},"lastHeartbeatAt":null,"metadata":null,"createdAt":"2026-03-14T06:09:38.711Z","updatedAt":"2026-03-14T07:30:02.678Z","urlKey":"app-store-optimizer","chainOfCommand":[{"id":"1e9fc1f3-e016-40df-9d08-38289f90f2ee","name":"CEO","role":"ceo","title":null}]}
|
||||
@@ -1,95 +0,0 @@
|
||||
# FrenoCorp Product Alignment
|
||||
|
||||
**Date:** 2026-03-08
|
||||
**Participants:** CEO (1e9fc1f3), CTO (13842aab)
|
||||
**Status:** In Progress
|
||||
|
||||
---
|
||||
|
||||
## Current Asset
|
||||
|
||||
**AudiobookPipeline** - TTS-based audiobook generation system
|
||||
- Uses Qwen3-TTS 1.7B models for voice synthesis
|
||||
- Supports epub, pdf, mobi, html input formats
|
||||
- Features: dialogue detection, character voice differentiation, genre analysis
|
||||
- Output: WAV/MP3 at -23 LUFS (audiobook standard)
|
||||
- Tech stack: Python, PyTorch, MLX
|
||||
|
||||
---
|
||||
|
||||
## Key Questions for Alignment
|
||||
|
||||
### 1. Product Strategy
|
||||
|
||||
**Option A: Ship AudiobookPipeline as-is**
|
||||
- Immediate revenue potential from indie authors
|
||||
- Clear use case: convert books to audiobooks
|
||||
- Competition: existing TTS services (Descript, Play.ht)
|
||||
- Differentiation: character voices, multi-narrator support
|
||||
|
||||
**Option B: Pivot to adjacent opportunity**
|
||||
- Voice cloning for content creators?
|
||||
- Interactive fiction/audio games?
|
||||
- Educational content narration?
|
||||
|
||||
### 2. MVP Scope
|
||||
|
||||
**Core features for V1:**
|
||||
- [ ] Single-narrator audiobook generation
|
||||
- [ ] Basic character voice switching
|
||||
- [ ] epub input (most common format)
|
||||
- [ ] MP3 output (universal compatibility)
|
||||
- [ ] Simple CLI interface
|
||||
|
||||
**Nice-to-have (post-MVP):**
|
||||
- Multi-format support (pdf, mobi)
|
||||
- ML-based genre classification
|
||||
- Voice design/customization UI
|
||||
- Cloud API for non-technical users
|
||||
|
||||
### 3. Technical Decisions
|
||||
|
||||
**Infrastructure:**
|
||||
- Self-hosted vs cloud API?
|
||||
- GPU requirements: consumer GPU (RTX 3060+) vs cloud GPUs?
|
||||
- Batch processing vs real-time?
|
||||
|
||||
**Monetization:**
|
||||
- One-time purchase ($99-199)?
|
||||
- Subscription ($29-49/month)?
|
||||
- Pay-per-hour of audio?
|
||||
|
||||
### 4. Go-to-Market
|
||||
|
||||
**Target customers:**
|
||||
- Indie authors (self-publishing on Audible/Amazon)
|
||||
- Small publishers (budget constraints, need cost-effective solution)
|
||||
- Educational institutions (text-to-speech for accessibility)
|
||||
|
||||
**Distribution:**
|
||||
- Direct sales via website?
|
||||
- Marketplace (Gumroad, Etsy)?
|
||||
- Partnerships with publishing platforms?
|
||||
|
||||
---
|
||||
|
||||
## Next Steps
|
||||
|
||||
1. **CEO to decide:** Product direction (AudiobookPipeline vs pivot)
|
||||
2. **CTO to estimate:** Development timeline for MVP V1
|
||||
3. **Joint decision:** Pricing model and target customer segment
|
||||
4. **Action:** Create technical architecture document
|
||||
5. **Action:** Spin up Founding Engineer on MVP development
|
||||
|
||||
---
|
||||
|
||||
## Decisions Made Today
|
||||
|
||||
- Product: Continue with AudiobookPipeline (existing codebase, clear market)
|
||||
- Focus: Indie author market first (underserved, willing to pay for quality)
|
||||
- Pricing: Subscription model ($39/month for 10 hours of audio)
|
||||
- MVP deadline: 4 weeks
|
||||
|
||||
---
|
||||
|
||||
*Document lives at project root for cross-agent access. Update as alignment evolves.*
|
||||
@@ -1,462 +0,0 @@
|
||||
# Technical Architecture: AudiobookPipeline Web Platform
|
||||
|
||||
## Executive Summary
|
||||
|
||||
This document outlines the technical architecture for transforming the AudiobookPipeline CLI tool into a full-featured SaaS platform with web interface, user management, and cloud infrastructure.
|
||||
|
||||
**Target Stack:** SolidStart + Turso (SQLite) + S3-compatible storage
|
||||
|
||||
---
|
||||
|
||||
## Current State Assessment
|
||||
|
||||
### Existing Assets
|
||||
- **CLI Tool**: Mature Python pipeline with 8 stages (parser → analyzer → annotator → voices → segmentation → generation → assembly → validation)
|
||||
- **TTS Models**: Qwen3-TTS-12Hz-1.7B (VoiceDesign + Base models)
|
||||
- **Checkpoint System**: Resume capability for long-running jobs
|
||||
- **Config System**: YAML-based configuration with overrides
|
||||
- **Output Formats**: WAV + MP3 with loudness normalization
|
||||
|
||||
### Gaps to Address
|
||||
1. No user authentication or multi-tenancy
|
||||
2. No job queue or async processing
|
||||
3. No API layer for web clients
|
||||
4. No usage tracking or billing integration
|
||||
5. CLI-only UX (no dashboard, history, or file management)
|
||||
|
||||
---
|
||||
|
||||
## Architecture Overview
|
||||
|
||||
```
|
||||
┌─────────────────────────────────────────────────────────────┐
|
||||
│ Client Layer │
|
||||
│ ┌───────────┐ ┌───────────┐ ┌─────────────────────────┐ │
|
||||
│ │ Web │ │ CLI │ │ REST API (public) │ │
|
||||
│ │ App │ │ (enhanced)│ │ │ │
|
||||
│ │ (SolidStart)│ │ │ │ /api/jobs, /api/files │ │
|
||||
│ └───────────┘ └───────────┘ └─────────────────────────┘ │
|
||||
└─────────────────────────────────────────────────────────────┘
|
||||
│
|
||||
▼
|
||||
┌─────────────────────────────────────────────────────────────┐
|
||||
│ API Gateway Layer │
|
||||
│ ┌──────────────────────────────────────────────────────┐ │
|
||||
│ │ Next.js API Routes │ │
|
||||
│ │ - Auth middleware (Clerk or custom JWT) │ │
|
||||
│ │ - Rate limiting + quota enforcement │ │
|
||||
│ │ - Request validation (Zod) │ │
|
||||
│ └──────────────────────────────────────────────────────┘ │
|
||||
└─────────────────────────────────────────────────────────────┘
|
||||
│
|
||||
▼
|
||||
┌─────────────────────────────────────────────────────────────┐
|
||||
│ Service Layer │
|
||||
│ ┌──────────┐ ┌──────────┐ ┌──────────┐ ┌────────────┐ │
|
||||
│ │ Job │ │ File │ │ User │ │ Billing │ │
|
||||
│ │ Service │ │ Service │ │ Service │ │ Service │ │
|
||||
│ └──────────┘ └──────────┘ └──────────┘ └────────────┘ │
|
||||
└─────────────────────────────────────────────────────────────┘
|
||||
│
|
||||
┌─────────────┼─────────────┐
|
||||
▼ ▼ ▼
|
||||
┌───────────────┐ ┌──────────────┐ ┌──────────────┐
|
||||
│ Turso │ │ S3 │ │ GPU │
|
||||
│ (SQLite) │ │ (Storage) │ │ Workers │
|
||||
│ │ │ │ │ (TTS Jobs) │
|
||||
│ - Users │ │ - Uploads │ │ │
|
||||
│ - Jobs │ │ - Outputs │ │ - Qwen3-TTS │
|
||||
│ - Usage │ │ - Models │ │ - Assembly │
|
||||
│ - Subscriptions│ │ │ │ │
|
||||
└───────────────┘ └──────────────┘ └──────────────┘
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## Technology Decisions
|
||||
|
||||
### Frontend: SolidStart
|
||||
|
||||
**Why SolidStart?**
|
||||
- Lightweight, high-performance React alternative
|
||||
- Server-side rendering + static generation out of the box
|
||||
- Built-in API routes (reduces need for separate backend)
|
||||
- Excellent TypeScript support
|
||||
- Smaller bundle sizes than Next.js
|
||||
|
||||
**Key Packages:**
|
||||
```json
|
||||
{
|
||||
"solid-start": "^1.0.0",
|
||||
"solid-js": "^1.8.0",
|
||||
"@solidjs/router": "^0.14.0",
|
||||
"zod": "^3.22.0"
|
||||
}
|
||||
```
|
||||
|
||||
### Database: Turso (SQLite)
|
||||
|
||||
**Why Turso?**
|
||||
- Serverless SQLite with libSQL
|
||||
- Edge-compatible (runs anywhere)
|
||||
- Built-in replication and failover
|
||||
- Free tier: 1GB storage, 1M reads/day
|
||||
- Perfect for SaaS with <10k users
|
||||
|
||||
**Schema Design:**
|
||||
```sql
|
||||
-- Users and auth
|
||||
CREATE TABLE users (
|
||||
id TEXT PRIMARY KEY,
|
||||
email TEXT UNIQUE NOT NULL,
|
||||
stripe_customer_id TEXT,
|
||||
subscription_status TEXT DEFAULT 'free',
|
||||
credits INTEGER DEFAULT 0,
|
||||
created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
|
||||
);
|
||||
|
||||
-- Processing jobs
|
||||
CREATE TABLE jobs (
|
||||
id TEXT PRIMARY KEY,
|
||||
user_id TEXT REFERENCES users(id),
|
||||
status TEXT DEFAULT 'pending', -- pending, processing, completed, failed
|
||||
input_file_id TEXT,
|
||||
output_file_id TEXT,
|
||||
progress INTEGER DEFAULT 0,
|
||||
error_message TEXT,
|
||||
created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
|
||||
completed_at TIMESTAMP
|
||||
);
|
||||
|
||||
-- File metadata (not the files themselves)
|
||||
CREATE TABLE files (
|
||||
id TEXT PRIMARY KEY,
|
||||
user_id TEXT REFERENCES users(id),
|
||||
filename TEXT NOT NULL,
|
||||
s3_key TEXT UNIQUE NOT NULL,
|
||||
file_size INTEGER,
|
||||
mime_type TEXT,
|
||||
purpose TEXT, -- input, output, model
|
||||
created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
|
||||
);
|
||||
|
||||
-- Usage tracking for billing
|
||||
CREATE TABLE usage_events (
|
||||
id TEXT PRIMARY KEY,
|
||||
user_id TEXT REFERENCES users(id),
|
||||
job_id TEXT REFERENCES jobs(id),
|
||||
minutes_generated REAL,
|
||||
cost_cents INTEGER,
|
||||
created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
|
||||
);
|
||||
```
|
||||
|
||||
### Storage: S3-Compatible
|
||||
|
||||
**Why S3?**
|
||||
- Industry standard for file storage
|
||||
- Cheap (~$0.023/GB/month)
|
||||
- CDN integration (CloudFront)
|
||||
- Lifecycle policies for cleanup
|
||||
|
||||
**Use Cases:**
|
||||
- User uploads (input ebooks)
|
||||
- Generated audiobooks (output WAV/MP3)
|
||||
- Model checkpoints (Qwen3-TTS weights)
|
||||
- Processing logs
|
||||
|
||||
**Directory Structure:**
|
||||
```
|
||||
s3://audiobookpipeline-{env}/
|
||||
├── uploads/{user_id}/{timestamp}_{filename}
|
||||
├── outputs/{user_id}/{job_id}/
|
||||
│ ├── audiobook.wav
|
||||
│ ├── audiobook.mp3
|
||||
│ └── metadata.json
|
||||
├── models/
|
||||
│ ├── qwen3-tts-voicedesign/
|
||||
│ └── qwen3-tts-base/
|
||||
└── logs/{date}/{job_id}.log
|
||||
```
|
||||
|
||||
### GPU Workers: Serverless or Containerized
|
||||
|
||||
**Option A: AWS Lambda (with GPU via EKS)**
|
||||
- Pros: Auto-scaling, pay-per-use
|
||||
- Cons: Complex setup, cold starts
|
||||
|
||||
**Option B: RunPod / Lambda Labs**
|
||||
- Pros: GPU-optimized, simple API
|
||||
- Cons: Vendor lock-in
|
||||
|
||||
**Option C: Self-hosted on EC2 g4dn.xlarge**
|
||||
- Pros: Full control, predictable pricing (~$0.75/hr)
|
||||
- Cons: Manual scaling, always-on cost
|
||||
|
||||
**Recommendation:** Start with **Option C** (1-2 GPU instances) + job queue. Scale to serverless later.
|
||||
|
||||
---
|
||||
|
||||
## Core Components
|
||||
|
||||
### 1. Job Processing Pipeline
|
||||
|
||||
```python
|
||||
# services/job_processor.py
|
||||
class JobProcessor:
|
||||
"""Processes audiobook generation jobs."""
|
||||
|
||||
async def process_job(self, job_id: str) -> None:
|
||||
job = await self.db.get_job(job_id)
|
||||
|
||||
try:
|
||||
# Download input file from S3
|
||||
input_path = await self.file_service.download(job.input_file_id)
|
||||
|
||||
# Run pipeline stages with progress updates
|
||||
stages = [
|
||||
("parsing", self.parse_ebook),
|
||||
("analyzing", self.analyze_book),
|
||||
("segmenting", self.segment_text),
|
||||
("generating", self.generate_audio),
|
||||
("assembling", self.assemble_audiobook),
|
||||
]
|
||||
|
||||
for stage_name, stage_func in stages:
|
||||
await self.update_progress(job_id, stage_name)
|
||||
await stage_func(input_path, job.config)
|
||||
|
||||
# Upload output to S3
|
||||
output_file_id = await self.file_service.upload(
|
||||
job_id=job_id,
|
||||
files=["output.wav", "output.mp3"]
|
||||
)
|
||||
|
||||
await self.db.complete_job(job_id, output_file_id)
|
||||
|
||||
except Exception as e:
|
||||
await self.db.fail_job(job_id, str(e))
|
||||
raise
|
||||
```
|
||||
|
||||
### 2. API Routes (SolidStart)
|
||||
|
||||
```typescript
|
||||
// app/routes/api/jobs.ts
|
||||
export async function POST(event: RequestEvent) {
|
||||
const user = await requireAuth(event);
|
||||
|
||||
const body = await event.request.json();
|
||||
const schema = z.object({
|
||||
fileId: z.string(),
|
||||
config: z.object({
|
||||
voices: z.object({
|
||||
narrator: z.string().optional(),
|
||||
}),
|
||||
}).optional(),
|
||||
});
|
||||
|
||||
const { fileId, config } = schema.parse(body);
|
||||
|
||||
// Check quota
|
||||
const credits = await db.getUserCredits(user.id);
|
||||
if (credits < 1) {
|
||||
throw createError({
|
||||
status: 402,
|
||||
message: "Insufficient credits",
|
||||
});
|
||||
}
|
||||
|
||||
// Create job
|
||||
const job = await db.createJob({
|
||||
userId: user.id,
|
||||
inputFileId: fileId,
|
||||
config,
|
||||
});
|
||||
|
||||
// Queue for processing
|
||||
await jobQueue.add("process-audiobook", { jobId: job.id });
|
||||
|
||||
return event.json({ job });
|
||||
}
|
||||
```
|
||||
|
||||
### 3. Dashboard UI
|
||||
|
||||
```tsx
|
||||
// app/routes/dashboard.tsx
|
||||
export default function Dashboard() {
|
||||
const user = useUser();
|
||||
const jobs = useQuery(() => fetch(`/api/jobs?userId=${user.id}`));
|
||||
|
||||
return (
|
||||
<div class="dashboard">
|
||||
<h1>Audiobook Pipeline</h1>
|
||||
|
||||
<StatsCard
|
||||
credits={user.credits}
|
||||
booksGenerated={jobs.data.length}
|
||||
/>
|
||||
|
||||
<UploadButton />
|
||||
|
||||
<JobList jobs={jobs.data} />
|
||||
</div>
|
||||
);
|
||||
}
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## Security Considerations
|
||||
|
||||
### Authentication
|
||||
- **Option 1:** Clerk (fastest to implement, $0-25/mo)
|
||||
- **Option 2:** Custom JWT with email magic links
|
||||
- **Recommendation:** Clerk for MVP
|
||||
|
||||
### Authorization
|
||||
- Row-level security in Turso queries
|
||||
- S3 pre-signed URLs with expiration
|
||||
- API rate limiting per user
|
||||
|
||||
### Data Isolation
|
||||
- All S3 keys include `user_id` prefix
|
||||
- Database queries always filter by `user_id`
|
||||
- GPU workers validate job ownership
|
||||
|
||||
---
|
||||
|
||||
## Deployment Architecture
|
||||
|
||||
### Development
|
||||
```bash
|
||||
# Local setup
|
||||
npm run dev # SolidStart dev server
|
||||
turso dev # Local SQLite
|
||||
minio # Local S3-compatible storage
|
||||
```
|
||||
|
||||
### Production (Vercel + Turso)
|
||||
```
|
||||
┌─────────────┐ ┌──────────────┐ ┌──────────┐
|
||||
│ Vercel │────▶│ Turso │ │ S3 │
|
||||
│ (SolidStart)│ │ (Database) │ │(Storage) │
|
||||
└─────────────┘ └──────────────┘ └──────────┘
|
||||
│
|
||||
▼
|
||||
┌─────────────┐
|
||||
│ GPU Fleet │
|
||||
│ (Workers) │
|
||||
└─────────────┘
|
||||
```
|
||||
|
||||
### CI/CD Pipeline
|
||||
```yaml
|
||||
# .github/workflows/deploy.yml
|
||||
name: Deploy
|
||||
on:
|
||||
push:
|
||||
branches: [main]
|
||||
|
||||
jobs:
|
||||
test:
|
||||
runs-on: ubuntu-latest
|
||||
steps:
|
||||
- uses: actions/checkout@v4
|
||||
- run: npm ci
|
||||
- run: npm test
|
||||
|
||||
deploy:
|
||||
needs: test
|
||||
runs-on: ubuntu-latest
|
||||
steps:
|
||||
- uses: vercel/actions@v2
|
||||
with:
|
||||
token: ${{ secrets.VERCEL_TOKEN }}
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## MVP Implementation Plan
|
||||
|
||||
### Phase 1: Foundation (Week 1-2)
|
||||
- [ ] Set up SolidStart project structure
|
||||
- [ ] Integrate Turso database
|
||||
- [ ] Implement user auth (Clerk)
|
||||
- [ ] Create file upload endpoint (S3)
|
||||
- [ ] Build basic dashboard UI
|
||||
|
||||
### Phase 2: Pipeline Integration (Week 2-3)
|
||||
- [ ] Containerize existing Python pipeline
|
||||
- [ ] Set up job queue (BullMQ or Redis)
|
||||
- [ ] Implement job processor service
|
||||
- [ ] Add progress tracking API
|
||||
- [ ] Connect GPU workers
|
||||
|
||||
### Phase 3: User Experience (Week 3-4)
|
||||
- [ ] Job history UI with status indicators
|
||||
- [ ] Audio player for preview/download
|
||||
- [ ] Usage dashboard + credit system
|
||||
- [ ] Stripe integration for payments
|
||||
- [ ] Email notifications on job completion
|
||||
|
||||
---
|
||||
|
||||
## Cost Analysis
|
||||
|
||||
### Infrastructure Costs (Monthly)
|
||||
|
||||
| Component | Tier | Cost |
|
||||
|-----------|------|------|
|
||||
| Vercel | Pro | $20/mo |
|
||||
| Turso | Free tier | $0/mo (<1M reads/day) |
|
||||
| S3 Storage | 1TB | $23/mo |
|
||||
| GPU (g4dn.xlarge) | 730 hrs/mo | $548/mo |
|
||||
| Redis (job queue) | Hobby | $9/mo |
|
||||
| **Total** | | **~$600/mo** |
|
||||
|
||||
### Unit Economics
|
||||
|
||||
- GPU cost per hour: $0.75
|
||||
- Average book processing time: 2 hours (30k words)
|
||||
- Cost per book: ~$1.50 (GPU only)
|
||||
- Price per book: $39/mo subscription (unlimited, but fair use)
|
||||
- **Gross margin: >95%**
|
||||
|
||||
---
|
||||
|
||||
## Next Steps
|
||||
|
||||
1. **Immediate:** Set up SolidStart + Turso scaffolding
|
||||
2. **This Week:** Implement auth + file upload
|
||||
3. **Next Week:** Containerize Python pipeline + job queue
|
||||
4. **Week 3:** Dashboard UI + Stripe integration
|
||||
|
||||
---
|
||||
|
||||
## Appendix: Environment Variables
|
||||
|
||||
```bash
|
||||
# Database
|
||||
TURSO_DATABASE_URL="libsql://frenocorp.turso.io"
|
||||
TURSO_AUTH_TOKEN="..."
|
||||
|
||||
# Storage
|
||||
AWS_ACCESS_KEY_ID="..."
|
||||
AWS_SECRET_ACCESS_KEY="..."
|
||||
AWS_S3_BUCKET="audiobookpipeline-prod"
|
||||
AWS_REGION="us-east-1"
|
||||
|
||||
# Auth
|
||||
CLERK_SECRET_KEY="..."
|
||||
NEXT_PUBLIC_CLERK_PUBLISHABLE_KEY="..."
|
||||
|
||||
# Billing
|
||||
STRIPE_SECRET_KEY="..."
|
||||
STRIPE_WEBHOOK_SECRET="..."
|
||||
|
||||
# GPU Workers
|
||||
GPU_WORKER_ENDPOINT="https://workers.audiobookpipeline.com"
|
||||
GPU_API_KEY="..."
|
||||
```
|
||||
@@ -1,196 +0,0 @@
|
||||
# Technical Architecture Document
|
||||
|
||||
**Date:** 2026-03-08
|
||||
**Version:** 1.0
|
||||
**Author:** CTO (13842aab)
|
||||
**Status:** Draft
|
||||
|
||||
---
|
||||
|
||||
## Executive Summary
|
||||
|
||||
AudiobookPipeline is a TTS-based audiobook generation system using Qwen3-TTS 1.7B models. The architecture prioritizes quality narration with character differentiation while maintaining reasonable GPU requirements for indie author use cases.
|
||||
|
||||
---
|
||||
|
||||
## System Architecture
|
||||
|
||||
```
|
||||
┌─────────────────┐ ┌──────────────────┐ ┌─────────────────┐
|
||||
│ Client App │────▶│ API Gateway │────▶│ Worker Pool │
|
||||
│ (CLI/Web) │ │ (FastAPI) │ │ (GPU Workers) │
|
||||
└─────────────────┘ └──────────────────┘ └─────────────────┘
|
||||
│ │
|
||||
▼ ▼
|
||||
┌──────────────┐ ┌──────────────┐
|
||||
│ Queue │ │ Models │
|
||||
│ (Redis) │ │ (Qwen3-TTS) │
|
||||
└──────────────┘ └──────────────┘
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## Core Components
|
||||
|
||||
### 1. Input Processing Layer
|
||||
|
||||
**Parsers Module**
|
||||
- epub parser (primary format - 80% of indie books)
|
||||
- pdf parser (secondary, OCR-dependent)
|
||||
- html parser (for web-published books)
|
||||
- mobi parser (legacy support)
|
||||
|
||||
**Features:**
|
||||
- Text normalization and whitespace cleanup
|
||||
- Chapter/section detection
|
||||
- Dialogue annotation (confidence threshold: 0.7)
|
||||
- Character identification from dialogue tags
|
||||
|
||||
### 2. Analysis Layer
|
||||
|
||||
**Analyzer Module**
|
||||
- Genre detection (optional ML-based, currently heuristic)
|
||||
- Tone/style analysis for voice selection
|
||||
- Length estimation for batching
|
||||
|
||||
**Annotator Module**
|
||||
- Dialogue confidence scoring
|
||||
- Speaker attribution
|
||||
- Pacing markers
|
||||
|
||||
### 3. Voice Generation Layer
|
||||
|
||||
**Generation Module**
|
||||
- Qwen3-TTS 1.7B Base model (primary)
|
||||
- Qwen3-TTS 1.7B VoiceDesign model (custom voices)
|
||||
- Batch processing optimization
|
||||
- Retry logic with exponential backoff (5s, 15s, 45s)
|
||||
|
||||
**Voice Management:**
|
||||
- Narrator voice (auto-inferred or user-selected)
|
||||
- Character voices (diverse defaults to avoid similarity)
|
||||
- Voice cloning via prompt extraction
|
||||
|
||||
### 4. Assembly Layer
|
||||
|
||||
**Assembly Module**
|
||||
- Audio segment stitching
|
||||
- Speaker transition padding: 0.4s
|
||||
- Paragraph padding: 0.2s
|
||||
- Loudness normalization to -23 LUFS
|
||||
- Output format generation (WAV, MP3 @ 128kbps)
|
||||
|
||||
### 5. Validation Layer
|
||||
|
||||
**Validation Module**
|
||||
- Audio energy threshold: -60dB
|
||||
- Loudness tolerance: ±3 LUFS
|
||||
- Strict mode flag for CI/CD
|
||||
|
||||
---
|
||||
|
||||
## Technology Stack
|
||||
|
||||
### Core Framework
|
||||
- **Language:** Python 3.11+
|
||||
- **ML Framework:** PyTorch 2.0+
|
||||
- **Audio Processing:** SoundFile, librosa
|
||||
- **Web API:** FastAPI + Uvicorn
|
||||
- **Queue:** Redis (for async processing)
|
||||
|
||||
### Infrastructure
|
||||
- **GPU Requirements:** RTX 3060 12GB minimum, RTX 4090 recommended
|
||||
- **Memory:** 32GB RAM minimum
|
||||
- **Storage:** 50GB SSD for model weights and cache
|
||||
|
||||
### Dependencies
|
||||
```yaml
|
||||
torch: ">=2.0.0"
|
||||
soundfile: ">=0.12.0"
|
||||
librosa: ">=0.10.0"
|
||||
fastapi: ">=0.104.0"
|
||||
uvicorn: ">=0.24.0"
|
||||
redis: ">=5.0.0"
|
||||
pydub: ">=0.25.0"
|
||||
ebooklib: ">=0.18"
|
||||
pypdf: ">=3.0.0"
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## Data Flow
|
||||
|
||||
1. **Upload:** User uploads epub via CLI or web UI
|
||||
2. **Parse:** Text extraction with dialogue annotation
|
||||
3. **Analyze:** Genre detection, character identification
|
||||
4. **Queue:** Job added to Redis queue
|
||||
5. **Process:** GPU worker pulls job, generates audio segments
|
||||
6. **Assemble:** Stitch segments with padding, normalize loudness
|
||||
7. **Validate:** Check audio quality thresholds
|
||||
8. **Deliver:** MP3/WAV file to user
|
||||
|
||||
---
|
||||
|
||||
## Performance Targets
|
||||
|
||||
| Metric | Target | Notes |
|
||||
|--------|--------|-------|
|
||||
| Gen speed | 0.5x real-time | RTX 4090, batch=4 |
|
||||
| Quality | -23 LUFS ±1dB | Audiobook standard |
|
||||
| Latency | <5 min per chapter | For 20k words |
|
||||
| Concurrent users | 10 | With 4 GPU workers |
|
||||
|
||||
---
|
||||
|
||||
## Scalability Considerations
|
||||
|
||||
### Phase 1 (MVP - Week 1-4)
|
||||
- Single-machine deployment
|
||||
- CLI-only interface
|
||||
- Local queue (in-memory)
|
||||
- Manual GPU provisioning
|
||||
|
||||
### Phase 2 (Beta - Week 5-8)
|
||||
- FastAPI web interface
|
||||
- Redis queue for async jobs
|
||||
- Docker containerization
|
||||
- Cloud GPU option (RunPod, Lambda Labs)
|
||||
|
||||
### Phase 3 (Production - Quarter 2)
|
||||
- Kubernetes cluster
|
||||
- Auto-scaling GPU workers
|
||||
- Multi-region deployment
|
||||
- CDN for file delivery
|
||||
|
||||
---
|
||||
|
||||
## Security Considerations
|
||||
|
||||
- User audio files stored encrypted at rest
|
||||
- API authentication via API keys
|
||||
- Rate limiting: 100 requests/hour per tier
|
||||
- No third-party data sharing
|
||||
|
||||
---
|
||||
|
||||
## Risks & Mitigations
|
||||
|
||||
| Risk | Impact | Mitigation |
|
||||
|------|--------|------------|
|
||||
| GPU availability | High | Cloud GPU partnerships, queue-based scaling |
|
||||
| Model quality variance | Medium | Human review workflow for premium tier |
|
||||
| Format parsing edge cases | Low | Extensive test suite, graceful degradation |
|
||||
| Competition from big players | Medium | Focus on indie author niche, character voices |
|
||||
|
||||
---
|
||||
|
||||
## Next Steps
|
||||
|
||||
1. **Week 1:** Set up development environment, create ADRs for key decisions
|
||||
2. **Week 2-3:** Implement MVP features (single-narrator, epub, MP3)
|
||||
3. **Week 4:** Beta testing with 5-10 indie authors
|
||||
4. **Week 5+:** Character voice refinement, web UI
|
||||
|
||||
---
|
||||
|
||||
*Document lives at project root for cross-agent access. Update with ADRs as decisions evolve.*
|
||||
Reference in New Issue
Block a user