Watch LangExtract automatically discover extraction patterns from your data
No manual YAML schemas required. LangExtract analyzes your text and suggests extraction patterns automatically.
Adapts extraction patterns based on text type - medical notes get different schemas than scientific papers.
From sample text to structured data in seconds. No domain expertise or manual configuration needed.
# Manual schema.yaml - hours of work
extraction_schema:
patient_info:
age: "Extract age near 'year-old'"
gender: "Extract male or female"
medications:
drug: "Extract medication names"
dosage: "Extract dosage amounts"
# ... 50+ more lines of configuration
# AI discovers schema automatically
result = lx.extract(
text=medical_text,
prompt="Extract patient info for analytics",
examples=[simple_example]
)
# ✅ Complete schema discovered automatically
# ✅ No manual configuration required
Automatically discovered categories:
Completely different patterns discovered:
Interactive network graphs showing relationships between extracted entities across both domains.
Left: Medical domain relationships (patient data, conditions, treatments)
Right: Scientific domain relationships (methods, performance, datasets)
Ready to explore AI-powered schema discovery? Get started in minutes.
git clone https://github.com/knightsri/aperio cd aperio python -m venv aperio_env source aperio_env/Scripts/activate pip install -r requirements.txt cp .env.example .env # Add your Gemini API key to .env jupyter notebook aperio_demo.ipynb