The Indico IPA Platform scales to meet the enterprise demands of unstructured data.
The Indico IPA Platform provides paper-driven organizations an AI- based technology that ushers the user through a technology-driven document processing workflow. The intuitive design allows those that are hands-on with the processing to easily set up models and ingest documents to initially train the model with the result of accelerating the production flow.
Indico IPA Product Modules
Point and click creation of AI powered custom models for the analysis, classification and comparison of unstructured content.
Document Ingestion & Workflows
The industry’s only solution covering structured, semi-structured and unstructured document formats with document ingestion designed for enterprise scale.
Flexible Document Ingestion
- GraphQL API for powerful document ingestion capabilities scaling to millions of documents per month
- TIFF, PDF (native and scanned), DOCX, JPG and more
Integrated Optical Character Recognition
- Integrated optical character recognition with built in quality and error correction to maximize output quality.
Custom Document Models
Custom machine learning models that understand your documents with a fraction of training data compared to traditional machine learning.
Visual Document Labeling Interface
- Keeping the context of a document look and feel while allowing model training inputs to be as easy as highlighting text.
Custom Document Attributes
- In three easy steps, your model is created and you’re ready to start training with the precise document attributes you define.
Machine Assisted Labeling
- With as few as 50 labeled documents your custom model comes to life with predictions appearing to indicate that the model is learning from your inputs.
Award-winning AI Explainability
Practical AI Explainability to deliver complete transparency and confidence in your custom models.
Automatic 80/20 Train/Test Split
- Automatic generation of a random 80/20 train/test split of your data to generate a complete and reliable set of machine learning performance metrics.
Data Science Best Practice Model Metrics
- F1, Precision, Recall and 4 different span types to examine to identify areas within the model for performance improvement.
Complete Traceability Of All Predictions
- Explore your model’s training data to find and correct errant labels to improve models performance in just a matter of clicks.
Support For The Simple To The Complex With Workflow Orchestration
Indico Workflows allows you to see all of the touchpoints of your documents from unstructured input to structured output. Create, manage, and modify your models and processors from one central location.
“Low Code” Buildable Canvas
- Visual workflow canvas to create and configure multi-step processes with multiple custom models .
Settings & Configurations
- Apply and modify settings for your Workflow to ensure that your output is precise to your process specifications.
Metrics & Analytics
- Understand the performance of your Workflow, from document processing time to time on task for your human in the loop Review.
Output Review & Validation
Your SME’s “bionic arm” with full control over your structured output with point and click intuitive document review and validation.
- Accept, reject, correct or insert missing values right on the document itself. Output data is “anchored” to its location in the document enabling rapid navigation between fields and pages.
Customizations for Every Scenario
- Create custom settings per Workflow to maximize optimization, including straight through processing by field or document, 2- or 4-eye review, and quality control oversampling among others.
- Designate SLA per Workflow and access queues to review backlog. Exceptions queues can be configured for manager oversight for edge cases or rejected documents.
APIs & Integrations
Indico’s GraphQL API delivers the flexibility to meet your unique implementation requirements. Leverage out of the box partners or create custom integrations for data ingest and output.
- Flexible GraphQL API accessible via client libraries in C#, Python, and Java, and with API access directly within the application.
- Support for JSON, XML and CSV outputs. Robust structured payload per item, including class values, confidence scores, positional detail and more. Customize your output for your downstream system needs; OCR text and metadata, formats, and fields are all configurable.
Out of the Box Integrations
- Pre-built integrations modules with popular enterprise platforms such as UiPath, Automation Anywhere, BluePrism and more.
Robust User Management and Security
- Integrated support for enterprise single sign on with detailed user roles and permissioning for maximum data protection.