In this short walkthrough, you will meet Herbie, our AI actuary inside Swallow

Starting from a single prompt, Herbie designs a complete insurance pricing model for electric vehicles in Germany, including inputs, outputs, decision rules and risk factors. Y

ou will see how the model is configured in Build, how tests and logs give full transparency over every step in the premium calculation, and how the final model is deployed as an API with live data streaming into your analytics. The transcript below captures the exact flow from first prompt to live, scalable pricing model.

‍

Herbie introduces the process

This screen opens with Herbie, your AI Actuary, introducing how Swallow helps you create complex insurance pricing models in minutes. The experience begins with a simple idea that you can describe in natural language, ready for Herbie to turn into a structured rating model.

‍

Setting the modelling goal

Here the video sets the scenario: creating an electric vehicle model for Germany that adjusts pricing based on mileage. Herbie takes the full prompt and prepares to generate every component needed for a complete pricing model.

‍

Generating the full model from your prompt

This step shows how your prompt is transformed into a structured model, complete with inputs, outputs, decisions, and risk pricing factors. Herbie prepares each part of the rating logic, giving you a clear foundation to refine inside Swallow.

‍

Reviewing the generated steps

Once the model has been created, Herbie presents a detailed written breakdown of every step. This gives you a high level summary of the structure and logic before diving into the platform itself to explore each part.

‍

Entering Swallow to explore the model

Now we arrive inside Swallow. The generated project is ready to explore, test, and configure.

This part highlights how easily you can run tests to see how the new model behaves. You can check premium outcomes, adjust scenarios, and refine logic until everything performs exactly as intended.

‍

Editing steps inside Build

Here the video shows the Build area, where every step of the model can be configured. You can group logic, create rules, apply loadings and define rating factors. Inputs and outputs are set up here to shape how the model receives and returns data.

‍

Inspecting logs for full transparency

This step opens the Logs view, giving you a transparent, line by line breakdown of how the premium was calculated. Every factor, formula and value is shown clearly, supporting debugging, audit, and model validation.

‍

Publishing and deployment options

Once the model is ready, you can deploy it in seconds. Swallow lets you serve the model directly to your customers through your connected systems using powerful and reliable APIs.

‍

Viewing real time customer data

This view shows how live customer data flows into the platform as your model is used in production. You can analyse volume, pricing behaviour and performance metrics in real time, all inside the Data area.

Here you see a snapshot of the live performance dashboard. With all user interactions captured instantly, you can track trends, spot anomalies, and refine your models with confidence.

‍

Book a demo to get started

The final step invites viewers to learn more and start a free trial. You can schedule a call to see how Swallow fits your organisation and begin building your own pricing ecosystem.

‍