Riders, parsed.
Upload a DJ or band technical rider PDF. Get structured production data — stage plots, input lists, equipment requirements — ready for your festival workflow.
The rider problem
Every festival production manager knows the pain: dozens of artists, each with a PDF technical rider in a different format. Stage dimensions, input lists, equipment requirements, power needs — all buried in unstructured documents.
Someone has to read each rider, extract the relevant data, and enter it into spreadsheets manually. It takes hours, introduces errors, and has to be repeated every event.
From PDF to production data
TRACE reads technical rider PDFs and extracts structured data automatically. Stage plot dimensions, input channel lists, equipment requirements, power specifications, and monitoring preferences — all parsed into clean, machine-readable formats.
The extraction uses local AI to understand document structure regardless of format. No two riders look alike, but TRACE handles the variation.
See it in action
Architecture
TRACE runs as a cloud-native pipeline. The React frontend on Vercel handles uploads and the festival dashboard. FastAPI on Cloud Run manages parsing orchestration and business logic.
Claude API extracts structured data from PDF riders — stage plots, input lists, equipment requirements, crew assignments. Cloud Tasks handles async batch processing.
All data lives in PostgreSQL via Supabase. PDFs are stored in Google Cloud Storage. Stripe handles per-festival billing.
What TRACE includes
Everything a production team needs to process technical riders at scale.
PDF parsing
Reads any technical rider PDF — text-based, scanned, or mixed. Handles tables, diagrams, and free-form text.
Stage plot extraction
Identifies stage dimensions, positions, and layout from rider diagrams and descriptions.
Input list parsing
Extracts channel-by-channel input lists with instrument, microphone, and DI requirements.
Equipment requirements
Pulls backline, monitoring, and power requirements into structured fields.
Batch processing
Process an entire festival lineup in one run. Dozens of riders, minutes instead of hours.
Multiple export formats
JSON, CSV, and production-specific formats. Ready for your existing workflow tools.
Local AI processing
All extraction runs locally. No rider data leaves your machine.
Stack
Built for reliability and local-first operation.
- Runtime
- Python
- AI
- Local LLM (Ollama)
- PDF Parsing
- pdfplumber + custom extractors
- Export
- JSON, CSV
- API
- FastAPI
- License
- AGPL-3.0
Get started
Clone, install, and run. Requires Python 3.11+ and Ollama.
# Clone and run TRACE
git clone https://github.com/formray/trace.git
cd trace
pip install -r requirements.txt
# Start Ollama (required for AI extraction)
ollama pull llama3
# Parse a technical rider
python -m trace parse rider.pdf --output jsonPricing
TRACE Core is free and open source. Support tiers for production companies and festivals.
Community
Free
Individual use
Pro
Contact us
Production companies
Enterprise
Contact us
Festivals and venues
All prices per organization. Annual billing.
The name
TRACE — TRACE — Technical Rider Automated Content Extraction.
Like tracing a signal path through a mixing console, TRACE follows the structure of a technical rider and extracts what matters. From document to data, precisely.
Open source
AGPL-3.0 licensed. The complete extraction engine is free and open source. Use it, improve it, contribute back.