How does the ingestion process work?
You can ingest documents by uploading them manually or through automated processes. The system processes documents at the folder level, meaning all documents within a folder are processed together when you execute a workflow.
Can I process different types of documents?
Yes, the system can process various document types, including leases, invoices, text documents, case laws, contracts, and emails.
How do I know which documents have been processed?
The system provides logs that indicate which documents have been started, indexed, and completed. In the near future, there will be a visual indicator (e.g., a checkmark) next to the file name to show its status.
Is the system able to handle large volumes of documents?
Yes, the system is designed to handle both small and large volumes of documents, scaling to millions of documents if needed.
How are documents with specific structures like contracts, purchase orders, and invoices handled?
For documents with structured information, like contracts, the system uses schema-based processing. It automatically identifies essential attributes and populates the values, making it easier to extract and analyze relevant information.
Can I have separate schemas for different types of documents?
Yes, you can create unique schemas for each type of document or folder, ensuring that the system accurately extracts and represents information specific to each document type.
Is there a chatbot interface for searching information?
Yes, the system has a chatbot interface that allows you to ask questions and get answers directly from your uploaded documents.
Can the system be customized or extended?
Yes, the system is a flexible platform that can be customized and extended to meet specific requirements and integrate with other applications.
What underlying technologies and components are used in the system?
The system utilizes Google Document AI, Vertex AI, Elasticsearch, Alloy DB, and other technologies to provide document processing, search capabilities, and database management.
Can the system generate tables or other formatted outputs based on queries?
The system can extract relevant information, and while it may not natively generate tables, customization and additional development can be done to enable table creation.
Does the system have the capability to highlight or annotate text within documents to show where answers are found?
While the current interface may not have this feature, it can be added through customization. The underlying technology (Gemini) does have the capability to perform contextual searches and highlight relevant text.
What is the pricing structure for the Trillo Doc AI solution?
The pricing includes a startup fee for customization and a recurring monthly licensing fee. The Google Cloud costs are separate and depend on the volume of data and usage. The recurring cost for the Trillo Doc AI application layer starts at $800 per month for a two-node cluster, and scales in increments of $800 per month with additional cluster nodes.
What is included in the startup fee?
The startup fee covers the initial customization of the system, including tailoring it to your specific requirements, loading documents, and ensuring correct indexing and setup.
Are there any additional costs for support or future customizations?
Basic support for application issues is included. However, additional customizations or feature requests may incur additional charges.