Telligro

Following the whole transaction

Telligro’s Ubix platform coordinates transactions between organisations. Work can arrive through email, forms, files, messages or portals. The platform follows the required steps, updates systems and brings people in for judgment or exceptions.

Its public examples include tax, insurance and lending workflows. Telligro describes a shared execution platform configured for those processes, with a record of actions and approvals.

Why this is technically difficult

A transaction is more than a document moving from A to B. It has a state: waiting for information, ready for approval, submitted or completed. Different organisations may hold different versions of that state.

I would test an email arriving twice, a changed attachment and a portal timing out halfway through submission. The system needs to know what already happened. Idempotency matters here: retrying the same request should not create a second transaction.

I would also want a clear boundary between extracting information, deciding what it means and taking an action. A person should be able to review an uncertain detail without restarting the entire workflow.

The business model and moat

The buyer is paying to remove coordination work and avoid mistakes. I would evaluate the cost per completed transaction, including the cases that still need manual attention. A fast first step is less valuable if the work gets stuck again later.

The potential moat is knowledge of how specific transactions really work: their exceptions, approvals and connections. Reusable handling of those details could make the next customer easier to serve.

There is an important balance here. Deep industry knowledge helps, but too much custom work can make each deployment a separate project. I would look for a common engine underneath the specialised workflows, and evidence that changes can be made without rebuilding it.

My take

I find the work between organisations interesting because it is easy to overlook. Nobody owns the whole process, yet everyone feels the delays. A product that reliably closes those gaps could become useful infrastructure. I would start with one complete transaction type and expand from there.