Crata AI built a conversational AI assistant for Tecniseguros that guides customers through the claims process, automatically validates documents and generates responses based on the type of claim. The result: a 40 per cent faster claims process and a 25-point improvement in the NPS.
A manual process that held back both the customer and the team
When Tecniseguros approached Crata AI, claims processing was an entirely manual, paper-based process. Every claim that came in triggered a chain of steps that depended on people: reviewing the accompanying documentation, checking it against the requirements for that type of claim, identifying what was missing, and returning the case to the policyholder to complete the information. Each of these steps added hours – and in many cases days – to the total response time.
The cost was not just in terms of time. Manual validation introduced a margin for human error that was difficult to control: misfiled documents, overlooked requirements, and inconsistent responses depending on who was handling the case. Above all, the policyholder was left in the dark. They had no way of knowing what stage their claim was at or what was expected of them, at precisely the most sensitive moment in their entire relationship with their insurer.
For a sector where the speed and clarity of the response are the key factors in customer satisfaction, this process was not a minor operational detail: it was a competitive disadvantage. The opportunity was clear: to turn the most frustrating moment of the customer experience into the smoothest.
The team at Tecniseguros knew that the answer lay in automation, but they did not want a generic tool that would force their team to completely change the way they worked. They needed a solution that would adapt to their actual workflows, the nature of their claims and their regulations – not the other way round.
An AI assistant that guides the policyholder through the entire claims process
We designed and implemented a conversational assistant that guides the policyholder step by step from the very start. Instead of a static form, the customer engages in a natural conversation in which the system asks for exactly what it needs depending on the type of claim being made, and validates each document as it is uploaded.
At the heart of the system is a natural language understanding engine that interprets both what the customer types and the documentation they attach. When it detects that information is missing or that a document does not meet the requirements, it resolves the issue immediately, without waiting for a subsequent manual review. When a case is complex enough to require human judgement, it is transferred to an agent with all the context already gathered, so that the team does not have to start from scratch.
- Conversational chatbot for opening and tracking claims
- Automatic document validation upon upload
- Dynamic responses depending on the type of claim reported
- Real-time status visibility for the policyholder
- Transfer to a human agent with all the context when necessary
One design decision was key: the system was integrated with the workflows and tools that Tecniseguros was already using. The team did not have to migrate to a new platform or relearn how to do their jobs. The AI was incorporated as a layer that enhances the existing process, which meant the impact was felt from the very first weeks rather than after a long adoption period.
That approach – building around the client’s actual process rather than imposing a generic tool – is what made the difference between a pilot project that ends up gathering dust and a solution that the team truly embraced.
Heading 1
Heading 2
Heading 3
Heading 4
Heading 5
Heading 6
Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur.
Block quote
Ordered list
- Item 1
- Item 2
- Item 3
Unordered list
- Item A
- Item B
- Item C
Bold text
Emphasis
Superscript
Subscript