All solutions
ALIBORA IDP · INTELLIGENT DOCUMENT PROCESSING
ALIBORA IDP

Turn documents intostructured data.

Build your own document-processing project, configure the extraction workflow, deploy an API, upload a document and receive machine-readable JSON that your business process can use immediately.

Request an IDP demoExplore the API flow
Project-basedAPI-firstStructured JSONMulti-model AIBring your own model
Document → intelligence → JSON
Invoice Extraction API
Any business document
IDP workflow
Structured JSON
application/json
{
  "document_type": "invoice",
  "supplier": "ACME Services",
  "invoice_number": "INV-2026-00842",
  "date": "2026-08-10",
  "currency": "XOF",
  "total": 1485000,
  "confidence": 0.97
}

THE ENGINE

The model does the reading. Your team does the deciding.

IDP classifies each incoming document, locates the fields that matter and returns them as structured data, so people stop retyping invoices, forms and records and start working on the exceptions instead.

  • Classification & extraction
  • Confidence scoring
  • Human review queue
  • JSON over a simple API
AI processing symbol on a laptop screen representing automated document extraction
FROM DOCUMENTS TO AUTOMATION

Stop building one-off document parsers for every business process.

IDP gives organizations a reusable platform for turning semi-structured and unstructured documents into predictable data. Each use case becomes a configurable project that can be consumed through an API.

Create projects

Separate document-processing use cases into projects with their own prompts, profiles, files, workflows and deployment configuration.

Design workflows

Chain extraction and AI-processing steps so each project follows the logic required by the business process.

Deploy an API

Publish a configured workflow as an API endpoint and connect it to ERP, CRM, portals, mobile apps or internal services.

Receive JSON

Turn the processing result into structured data that downstream software can validate, store, route or automate.

API-FIRST BY DESIGN

Create once. Integrate everywhere.

The product flow is deliberately simple for developers and operations teams: configure the intelligence in IDP, then call it from any system that can send an HTTP request.

01

Create a project

Define the document use case and organize its extraction configuration.

02

Configure intelligence

Select extraction components, prompts, models and workflow steps for the project.

03

Deploy & generate access

Create an API deployment and manage the API keys authorized to call it.

04

Upload a document

Send the document to the public deployment endpoint using the assigned authentication token.

05

Process asynchronously

IDP launches the configured workflow and returns a job identifier while the document is processed.

06

Retrieve structured output

Check the job status and receive the workflow result as JSON for the calling application.

PROJECT-BASED INTELLIGENCE

A workspace for every document use case.

Invoice extraction, identity documents, forms, contracts, application files, certificates or sector-specific documents can each be managed as independent projects instead of hard-coded parsers.

Project files

Upload and manage representative files used to configure and validate a project.

Text extractors

Choose and manage text-extraction components as part of the processing pipeline.

Prompts

Create project-specific prompts that transform extracted content into the structure your application expects.

Profiles

Organize project-level configuration and reusable processing behavior.

Workflows

Compose multiple processing stages instead of relying on one monolithic model call.

Workflow steps

Control the sequence of extraction, transformation and AI steps in the pipeline.

MULTI-MODEL AI

Use the right model for the document, including your own.

IDP is not tied to one embedded model. The platform manages LLM configurations and cloud-model access while allowing custom model endpoints to be registered with their own connection and runtime settings.

Multiple AI models

Register and manage multiple LLM configurations instead of forcing every project through a single model.

Cloud model access

Use configured cloud AI services where they are appropriate for the workload.

Bring your own model

Point IDP to a custom model endpoint through a configurable base URL and model identifier.

Runtime controls

Configure API version, retries, timeout, output-token limits and reasoning behavior per model configuration.

BYOM
Your model. Your endpoint. Same IDP workflow.
Model nameinvoice-extractor-v3
Modelmy-org/document-llm
Base URLhttps://ai.example.org/v1
Timeout120 s
Max output tokens4096
Reasoning modeConfigurable
PRODUCTION API DEPLOYMENTS

Turn a workflow into a service your applications can call.

IDP includes managed API deployments, deployment keys, public execution endpoints and job-status retrieval. Your developers integrate one stable service while your AI workflow can evolve behind it.

API key management

Create, activate and manage keys attached to an API deployment.

Token-authenticated calls

Public processing endpoints validate the supplied x-auth-token before running a workflow.

Multipart file upload

Send a business document directly to the deployment endpoint for processing.

Asynchronous jobs

Receive a job ID immediately and poll the status endpoint for the processing result.

Replace logic without breaking consumers

Evolve project workflows while keeping downstream integrations centered on the deployment endpoint.

Integration-ready

Connect IDP with business applications, portals, back-office services and automated workflows.

API integration
One API call from your process
REQUEST
curl -X POST \
  https://idp.example/api/v1/deployments/api/acme/invoices \
  -H "x-auth-token: YOUR_API_KEY" \
  -F "file=@invoice.pdf"
RESPONSE
{
  "success": true,
  "message": "Task in progress",
  "data": {
    "job_id": "job_8f92..."
  }
}
COMPOSABLE PROCESSING PIPELINE

More than OCR. A complete document intelligence workflow.

Traditional OCR stops at text. IDP is designed to continue: extract content, apply AI reasoning, normalize fields and expose a structured result to another system.

Document

PDF · image · scan · business file

Extract

Text extraction / preprocessing

Workflow

Configured processing steps

AI models

Embedded, cloud or custom models

Structure

Normalized machine-readable output

Business system

ERP · CRM · BPM · portal · data platform

ORGANIZATION & GOVERNANCE

Built for shared organizational use, not isolated scripts.

The repository includes organization and user management alongside projects and deployments, making IDP a platform that multiple teams can operate rather than a developer-only extraction script.

Organizations & users

Manage organizations and attach users to the environments they are allowed to access.

Authenticated administration

Administrative project, workflow, model and deployment operations are protected by API authentication.

Project separation

Keep configurations and document-processing use cases organized as independent projects.

Controlled deployments

Deploy only approved configurations and control which API keys can invoke them.

USE CASES

One platform for documents across the organization.

Invoices & accounting

Extract supplier, invoice number, dates, tax, currency, line items and totals before sending data to finance systems.

Identity & onboarding

Extract identity-document fields and pass normalized data into enrollment, KYC or registration workflows.

Forms & applications

Turn application forms and administrative submissions into structured fields for case-management systems.

Contracts & legal documents

Identify parties, dates, clauses, references and metadata for document-management or approval processes.

Public administration

Digitize information from permits, certificates, declarations and legacy administrative files.

Sector-specific documents

Build project-specific extraction for logistics, healthcare, insurance, telecom, banking or other specialized documents.

FOR DEVELOPERS & OPERATIONS

Keep document intelligence behind a clean API boundary.

Business applications should not need to know which extractor, prompt or model produced the answer. IDP lets those details evolve independently behind a stable integration layer.

Version processing logic without rebuilding every consuming application
Centralize AI/model credentials and runtime configuration
Separate project design from API consumption
Use asynchronous processing for long-running document workloads
Return JSON that downstream applications can validate and persist
DOCUMENTS IN · STRUCTURED DATA OUT

Build document intelligence into every process.

Tell us what documents your organization receives and what data your systems need. ALIBORA IDP can turn that document flow into a reusable API-driven processing service.

Request an IDP demoDiscuss an integration