What sets Holistics apart from existing tools?

Holistics is differentiated by three key pillars.

Analytics As-Code: Bring DevOps Practices to BI

Data transformation tools and self-service BI often fall short in developer experience and robust metric modeling, leading to maintenance and reusability issues.

  • YAML/Jinja modeling is difficult to maintain.
  • The lack of robust metric modeling often leads to dependency on derived tables for complex operations, which makes it vulnerable to upstream logic changes (for example: dimension name change).
  • Popular visual BI and non-SQL BI tools, while feature-rich, are not designed with an ‘as-code’ approach. They either treat as-code as an after-thought (dumping JSON into Git) or struggle with reusability and accuracy of analytics definitions.

Solution: Analytics As Code

  • Define models and dashboards using code. Check in to Git version control. Perform code reviews before deploying.
  • Modeling language is well-designed (not string-based JSON/YAML/Jinja) with proper typing structure. This enables things like autocomplete, static typing and module systems, enhacing code reusability. Overall it ensures a higher level of developer productivity.

AMQL: Metrics As First-Class Citizens

AMQL is a metric-centric analytics language that provides a higher-level abstraction for defining metrics independently from data tables, which provide the following benefits:

  • Closer to business users’ mental models: People think about metrics, not dimensions and measures. AMQL allows building a self-serve interface that’s aligned with end-users’ natural way of asking business questions.
  • Empowers self-serve analytics: End user’s analytics interactions are translated into intermediary metric-centric queries before compiling into SQL.
  • Simplifies complex SQL functions, making advanced analytics more accessible. AMQL aligns with the business user mindset.

Analytics Canvas: Canvas-based Analytics Presentation

Problems: Traditional dashboards don’t provide a natural consumption experience to end-users’ way of thinking.

  • Linear flow, not natural to humans’ thinking.
  • Doesn’t provide enough business context.

Solution: Canvas-based analytics presentation engine. It allows analysts and business users to consume & explore data in the form that is more natural to human thinking process.