Skip to content
QUANDIS

QBO Platform · Agentic orchestration

QBO orchestrates AI. Not the other way around.

AI agents in QBO are steps in your workflow. QBO decides when a model runs, on what data, under whose security, and where the result goes, and results land as QBO data in your audit trail.

AI where you trust it. People where you need them. Your rules decide which.

A case’s workflow calls the model; the answer comes back to the case. Choose Claude, Gemini, OpenAI, or Amazon Bedrock by configuration.

What it does

Built into every module

AI methods on documents, scores, imports, forms, messages, reports, and queues — configured with templates, not custom code.

Documents

Classify, label, transcribe, and extract data with per-field and weighted confidence scores, including from original PDFs and images.

Records and forms

Turn a document into records — including nested ones — complete forms, and update fields from AI output.

Review and reply

Draft review memos and email replies for people to check before anything moves forward.

Plain-English reporting

Questions become a single read-only SELECT, validated against your schema without running it; drafts that fail validation come back flagged for review.

Scale and cost control

Provider batch processing through QBO’s queue, context caching where supported, and token costs posted to ledgers by type.

People in the loop, by rule

An AI agent can have its own user account and receive work through the same assignment rules as your people. Results carry a confidence score, so a worklist can send anything below your bar to a person for review, and worklist audits spot-check what the agent completed.

Live data in instructions

Agent instructions can pull live QBO data at run time, and QBO agents can use tools from external MCP servers.

Building blocks

AI enginesAI methods across documents, scores, imports, forms, messages, reports, and queues
Model providersChosen per agent by configuration

Integrations

  • Anthropic ClaudeClassification, extraction, review memos, replies, and batch processing.
  • Google GeminiThe same AI capabilities through Gemini on Google Cloud, with context caching for long instructions.
  • OpenAIThe same AI capabilities through OpenAI models.
  • Amazon BedrockThe same AI capabilities through models hosted in your AWS account.
  • MCP serversLet QBO’s AI agents use tools from external MCP servers.
All integrations →

More of the platform

See ai running on your own process.

Request a Demo