Jump to content
  • Spotfire Actions: From Insights to Impact

    Spotfire Actions facilitate swift, data-driven decision-making by leveraging extensive integrations across diverse verticals.


    Spotfire Actions make it possible to immediately act on insights. Through deep and secure integration, take action in more than 200 commercially-supported connections to various business systems, and more than 1800 community connectors and 100s of Accelerators. Actions bridge the gap between business insight and business decision by triggering seamless writing of transactions from your analytics platform to any database or application. Whether on-premise or in the cloud, even non-developers are enabled to build no-code business applications without the complexity of time-consuming integrations. Actions utilize not only a seamless integration with TCI, but are also available in any context that offers a REST API.

    In this article we will now cover three different examples of invoking an Action in Spotfire.

    ListenUp: Anomaly Detection on Manufacturing Audio


    In this example, we use a combination of TIBCO Data Science and Spotfire to detect anomalies in manufacturing audio data. Being a cheap alternative to expensive sensors, audio is a great way to quickly detect unexpected behavior. We train an ensemble of anomaly detection models (isolation forest, autoencoder, and local outlier factor) in TIBCO Data Science - Team Studio that is deployed on both batch and real-time data in Spotfire. When we find an unanticipated outlier, we use a Spotfire Action to appropriately notify the correct parties. The action sends both an email and/or Slack message to the person responsible for the machine where the anomaly occurred. This allows for an immediate response and action when things go wrong.

    Dynamic Pricing


    When observing models in production, it is valuable to highlight where there might be unexpected behavior. This use case utilizes models that predict whether or not a customer would renew their policy. In our analysis, we can key into specific instances that might be of interest to us; this could be where the model was very confident that a customer would renew, but they did not. What caused this? Was there anything in common between these customers? We use Spotfire Actions to help us investigate. Our Action, with the click of a button, sends out these instances to a Postgres database along with a user-prompted message. In this database, we develop a repository of cases to further investigate. 

    Pro Basketball Draft Analysis


    Spotfire Actions are a great tool to take snapshots of important insights from a dashboard. This analysis lets us compare and contrast basketball players who are up for selection in the NBA Draft. Interactively, we select and learn about each player, guided by advanced analytics and informative graphs. When we find a player we like, we use an Action that takes a screenshot of the page and saves it into a shared Google Drive folder. The insights that one person finds are democratized?now easily distributed to other decision-makers in the organization.

    Spotfire Actions stand as a revolutionary tool in the realm of business analytics, fostering an immediate response to insights. With its vast integration capabilities, ranging from more than 200 commercially-supported connections to a plethora of community connectors and Accelerators, it provides a swift bridge from insight to impactful decision-making. As showcased in the examples, this tool has diverse applications: detecting manufacturing anomalies using audio, aiding in dynamic pricing strategies, and assisting in the analysis of basketball players. Its adaptability to various contexts, whether on-premise or cloud-based, and its compatibility with REST API, empowers even non-developers to harness its power without delving into intricate integrations. In essence, Spotfire Actions embody the future of actionable analytics, facilitating swift, data-driven decisions across verticals.

    User Feedback

    Recommended Comments

    There are no comments to display.

  • Create New...