/

CONTACT US

The future of reconciliation is here

Learn more about Grath's managed reconciliation inference engine clearing the path for AI application builders in financial services.

/OFFICE

77 Coleman Street, London, EC2R 5BJ

/

FAQ

Frequently Asked Questions

What is the difference between Topa and Grath's reconciliation platform?

The Grath reconcilation platform is the application; Topa is the engine, model, and agent, exposed as infrastructure for teams building their own workflows.

Why not just use a general-purpose LLM?

General-purpose models aren't trained on reconciliation data, return unstructured text instead of match results, and send your ledgers to a horizontal inference provider.

Do I need to replace my current reconciliation tool?

Most customers start by pointing Topa at a specific problem area while their existing system continues to run.

How does pricing work?

You pay per token, metered separately for input and output, at rates that vary by layer.

Can I use Topa inside a regulated institution?

Yes, models are versioned and pinnable, and audit-grade logging is built in.

/

FAQ

Frequently Asked Questions

What is the difference between Topa and Grath's reconciliation platform?

The Grath reconcilation platform is the application; Topa is the engine, model, and agent, exposed as infrastructure for teams building their own workflows.

Why not just use a general-purpose LLM?

General-purpose models aren't trained on reconciliation data, return unstructured text instead of match results, and send your ledgers to a horizontal inference provider.

Do I need to replace my current reconciliation tool?

Most customers start by pointing Topa at a specific problem area while their existing system continues to run.

How does pricing work?

You pay per token, metered separately for input and output, at rates that vary by layer.

Can I use Topa inside a regulated institution?

Yes, models are versioned and pinnable, and audit-grade logging is built in.

/

FAQ

Frequently Asked Questions

What is the difference between Topa and Grath's reconciliation platform?

The Grath reconcilation platform is the application; Topa is the engine, model, and agent, exposed as infrastructure for teams building their own workflows.

Why not just use a general-purpose LLM?

General-purpose models aren't trained on reconciliation data, return unstructured text instead of match results, and send your ledgers to a horizontal inference provider.

Do I need to replace my current reconciliation tool?

Most customers start by pointing Topa at a specific problem area while their existing system continues to run.

How does pricing work?

You pay per token, metered separately for input and output, at rates that vary by layer.

Can I use Topa inside a regulated institution?

Yes, models are versioned and pinnable, and audit-grade logging is built in.