API Overview

PatternPulse AI exposes a RESTful API built on FastAPI. The API accepts PDF uploads, processes them through the ML pipeline, and returns classification results along with redacted documents.

Base URL:

ENDPOINThttps://api.patternpulse.ai/v1

Authentication: All API requests require a Bearer token in the Authorization header.

CURLcurl -X POST https://api.patternpulse.ai/v1/analyze \ -H "Authorization: Bearer YOUR_API_KEY" \ -F "file=@document.pdf"
All API responses are returned in JSON format. File downloads are returned as binary streams.

Supported File Types

FormatExtensionMax SizeStatus
PDF.pdf50 MBSupported
DOCX.docxComing Soon
DWG.dwgPlanned
DXF.dxfPlanned
Only PDF files are currently accepted. Uploading unsupported formats will return a 415 error.

Classification Categories

PatternPulse AI classifies construction documents into six standard engineering disciplines used across the AEC industry.

DisciplineCodeDescription
ArchitecturalARCHFloor plans, elevations, sections, interior layouts
StructuralSTRFoundation plans, framing, structural details
MechanicalMECHHVAC layouts, ductwork, mechanical schedules
PlumbingPLMBPiping layouts, fixture schedules, drainage
ElectricalELECPanel schedules, lighting plans, power layouts
Fire ProtectionFIRESprinkler layouts, fire alarm plans

Sensitive Information Detection

TypeMethodExamples
Person NamesspaCy NERJohn Doe, Sarah Lee
Email AddressesRegex + NERuser@example.com
Phone NumbersRegex(555) 123-4567
Physical AddressesspaCy NER123 Main Street, Suite 100
Project IDsRegexPRJ-2026-0042

Error Codes

CodeMessageDescription
400Bad RequestMissing or invalid parameters
401UnauthorizedInvalid or missing API key
413Payload Too LargeFile exceeds 50 MB
415Unsupported Media TypeFormat not supported
422Unprocessable EntityCannot extract text
429Rate LimitedToo many requests
500Internal Server ErrorServer-side error

Contact & Support

Have questions, feedback, or need help integrating PatternPulse AI? Reach out through any of the channels below.

Email

support@patternpulse.ai

GitHub

github.com/patternpulse

Documentation

docs.patternpulse.ai