Extract the information you need
Turn fields, tables, and marked options in PDFs and scanned documents into data your process can use. Combine rules and trainable models for each document type.
An agent platform for documents and data
Move the information in your documents into the work
Volerai processes documents and data, and can send the result to products, databases, and APIs.
Information should not stay trapped in documents
Documents arriving from different channels, data entered again and again, and disconnected review steps slow teams down. Volerai helps you turn the path from reading a document to sending the result into a defined process.
Shape the process around the work
Bring reading, review, decision, and transfer steps together in a visual workflow. Define the path that fits each document type.
Bring your team in where review is required
Inspect results next to the document, make the needed corrections, and continue with verified information.
One process from document to result
A Volerai flow combining reading, extraction, human review, and transfer.
- 1
Collect the documents
Upload files or take them in from configured folders and email channels. Keep related documents under the same job.
- 2
Read and classify
Turn the document into readable data. Use the model and processing path that fit the document type.
- 3
Extract and check
Define the fields you need. Check them with rules and route results that need review to your team.
- 4
Send it into the work
Move verified results to the relevant system through API, database, or document-management connections.
Flows, connections, and assistants
The no-code flow stays. A complex step can be added in Java or JavaScript. Integration is scoped to the target system, and more than one agent can work in the same process.
No-code flow
Design steps, conditions, human review, and transfer in a visual flow without writing code. The path each document takes stays visible on the flow.
See it in workflowsA Java or JavaScript step
The flow is designed without code. When a step is too complex for the visual pieces, that step can be written in Java or JavaScript as an extra path.
Straightforward integration
Move a verified result to an API, a database, an ERP, or another product you use. The connection is set up in the scenario with the target system.
See it in workflowsMultiple agents
More than one assistant can work in the same scenario. Each can use a different model, including a local model. Suggestions are reviewed. Writing or changing data stays locked until the scenario allows it.
See it in assistantsBuild a structure that learns your documents
Every organization has its own document layouts and information needs. In Volerai you define document types, fields, and training examples, then create classification and extraction models for your process. You can also label document fields and start training from a small set of examples.
Plan it around your infrastructure
Artificial intelligence models can be connected, and local models can be used within the installation. Which data goes to which model and which system is planned with you.
Label fields and train on a few examples
You mark fields on the document and turn them into training examples. Model training can start from a small set of examples. How many examples are enough is decided with the document type.
Add an assistant that works on the document
Create agents that explain where a value came from, help check fields, and offer corrections. Each agent can use a different model. Tools and permissions are set for the scenario. Actions are written to a detailed log, and roles plus who may use each agent can be defined.
Where did this amount come from?
The amount was read from the subtotal line. It was left for review so it can be compared with the VAT line.
Suggestion: the team confirms the field before any correction is applied.
Not applied
Logging, roles, and agent access
Actions are written to a detailed log. Roles separate viewing from running. You can define who may use each agent. Human approval can be part of the check step in the scenario. Writing data can stay locked until the scenario allows it.
Take the data, process it, send it on
Volerai is an agent-based product. It can take documents and data into a process, work on them, and move the result to the products, databases, or APIs you use.
Connect a model
Attach the artificial intelligence models the process needs.
Run locally
Local models can be used within the installation scope.
Choose a model per agent
Each agent can be defined with a different model.
Transfer the result
Verified output can go to an API, a database, or a product connection.
Let’s talk about your documents
Tell us which documents you process, which information you need, and where the result should go. We can look at how Volerai can be adapted to your process.
Request a demo