How Jedify Uses the Semantic Fusion™ Model

The Jedify platform leverages the Semantic Fusion™ Model through an agentic flow system. This process seamlessly transforms natural language questions into structured data responses:

Question Intake: User submits a natural language question
Profiler Agent: Analyzes the question using several components:

• Timeframe Analyzer: Identifies temporal aspects
• Question Classifier: Categorizes the query type
• Scope: Extracts the scope needed to answer the question

Entities Retrieval: Determines what information is required to answer
SQL Compilation: Transforms requirements into executable SQL leveraging the semantic model

Answer Composition: Assembles the various components of the response

• Data Normalizer: Ensures consistent formatting and units
• Output Generation: Produces charts, insights, and explanations

The Semantic Fusion™ Model serves as the infrastructure layer that enables this entire process, providing the semantic understanding necessary to interpret questions and compose appropriate SQL.