In March we’re introducing a new version of Xcall, an easier way to bring in calculations and data from other Spark services. Together with the Spark Assistant, you can now use Xcall as a native Excel formula, without using VBA or LAMBDA. The new Xcall formula =SPARK_XCALL() consists of four components, each of which builds on top of the previous one to help you construct your call. First, you'll be able to search through all the services you can access, then pick one and see all the available versions. Next, you'll be able to generate input and output templates for the service, from which you can choose only those inputs you want to change and only those outputs you wish to receive. Finally, when all four parameters are defined, Xcall will retrieve the requested data from the specified Spark service. This visual, iterative process is highly intuitive and eliminates the need to use arcane, nested formulas or learn another programming language!

Xcall enables common calculations to be standardized and centralized for use across an organization, eliminating local variations. Define the calculations as functions in one place, and reference them wherever they are needed. You can be confident in the knowledge that it’s always the correct version!

Xcall also offers flexibility when running tests or performing scenario analysis by enabling you to fix certain inputs and easily manipulate others. This shows how your calculations are impacted by changes in assumptions, empowering you to call on your shared functions to produce complex scenarios, simulations, and more!

For more information about setting up Spark Assistant and using the new Xcall, please consult our documentation. If you run into any issues or have questions, contact Support.

Model Comparison

Spark now offers the ability to compare any two versions of an Excel model. To see it in action: navigate to the Version Overview page, select two versions, click Compare, and Spark will generate a Model Comparison Report with a detailed analysis of the differences, clearly identifying what’s been changed. This greatly facilitates understanding of the model's development process and eliminates the need for tedious and time-consuming manual comparison. You also always have a precise and up-to-date record of the model’s change history, which is ideal for audits!

Webhooks

One of our customers was looking for a way to initiate an action with Spark whenever a service is uploaded. To address this requirement, we implemented a webhook for Spark!

Webhooks are a type of system-to-system communication following the Hollywood principle of “Don’t call us; we’ll call you!” One system sends a message to another system as soon as a specific event occurs, similar to a push notification on your smartphone. In the case of Spark, our webhook can pass information based on events such as: uploading a service, updating a service, running a testbed, and more. This introduces new ways to embed Spark capabilities more deeply into an operational process without having to develop a lot of additional code! Use cases range from simple actions such as developing notifications, assigning tasks to a workflow, or automating deployment!

If you are interested in having a webhook set up for your environment, please contact the Coherent team. We will be adding the ability to configure your own webhooks in a future release!

Integration improvements

With this release, we made it even easier to integrate your Spark service APIs. First, we added data object and response data object definitions to the Swagger definitions Spark provides. Swagger definitions are like blueprints for the APIs and can be used to:

• Easily import API calls into Postman and low-code tools.

• Create (and embed) interactive API documentation via Swagger UI or Redocly.

Second, all of our code snippets now include the Spark tenant name; there’s no need to look this up manually!

Bespoke user interfaces for every model upload

Spark Forms brings Spark’s power to your phone, your tablet, that old computer at the local library… almost any device! For every Excel model you upload, Spark can now generate a bespoke, responsive, web-based app. Share the link with colleagues (or display the QR code for them to scan) and get everyone testing and providing feedback sooner. You can even embed Spark Forms within your workbench or testing suite for convenience.

Use our web-based form editor to customize your input labels, reorder, or even remove them to create a custom testing interface. All without touching your original model. Reduce clutter and bring focus to the inputs (and outputs) that matter most.

Speak to your Customer Service representative about enabling Forms for your tenant and for a demo.

Other notable enhancements

• Improved ability to handle large CSV outputs by splitting them into multiple files.

• The API Call History functionality can now be disabled upon customer request.

• Improvements to Neuron around the DATEDIF functionality.

• Proper support for single row (1xN>1) vectors in the API and UI.

• Fixed Execute API requested_outputs to accommodate multiple filter items using either a string or JSON array.

Updated the Import Export tool to use .NET 6.0. It now works on Mac computers with ARM-based Apple silicon (M1 and M2 processors)!

Did you know?

We've been curious about how ChatGPT 4 from OpenAI can be used in combination with Excel but haven't seen many industry experts share any specifics. So, we went straight to the source! Here are the top four ways ChatGPT4 will be able to support Excel users, as shared by ChatGPT4 itself.

Should you have any questions about this release or interest in any of the updates above, please reach out to us via email.