Core Concepts

Fynite is built on a set of core concepts that define how data is integrated, structured, visualized, and acted upon. Understanding these concepts is essential for effectively using the platform.


Platform Architecture Overview

Fynite follows a layered architecture that separates responsibilities while enabling seamless interaction between components. Each layer plays a distinct role in transforming data into actionable outcomes.


Integration Layer

The Integration Layer is responsible for connecting Fynite to external systems and data sources.

It enables:

  • Data ingestion from APIs, databases, and third-party platforms

  • Synchronization of external data into Fynite

  • Enforcement of compliance and governance during data exchange

This layer ensures that all incoming and outgoing data is consistent, secure, and traceable.


Data Models (Master Data Management)

Data Models define how information is structured and managed within Fynite.

Key elements include:

  • Entities and attributes

  • Relationships between datasets

  • Centralized repositories of structured data

This layer provides a single source of truth, enabling consistent data usage across workflows, interfaces, and automation.


UI Management Layer (Fynite UI Studio)

The UI Management Layer enables the creation of dynamic, configurable user interfaces.

Using Fynite UI Studio, users can:

  • Build dynamic forms for data input and interaction

  • Configure UI behavior without writing code

  • Adapt interfaces based on workflows or data conditions

This layer allows teams to create tailored user experiences on top of structured data.


Visibility Layer

The Visibility Layer provides real-time insight into data and system activity.

It includes:

  • Dashboards for monitoring key metrics

  • Live views of data and workflows

  • Visualizations for analysis and decision-making

This layer ensures that users can observe and understand what is happening across the platform at any time.


AI Agents (Fynite Agent Studio)

AI Agents are configurable components that analyze data and make decisions.

They can:

  • Detect patterns and anomalies

  • Apply rules or models to data

  • Trigger actions or recommendations

AI Agents operate on top of structured data and are central to enabling intelligent automation within Fynite.


Execution Layer (Remediation Agents)

The Execution Layer is responsible for acting on insights generated by the system.

Remediation agents:

  • Execute workflows based on predefined conditions

  • Perform automated actions in response to AI decisions

  • Close the loop between detection and resolution

This layer ensures that insights lead to measurable outcomes.


Data Interaction and Relationships

Fynite enables users to work with data through structured interaction mechanisms.

This includes:

  • Data cataloging and organization

  • Defining relationships between datasets (join relationships)

  • Querying and exploring data

These capabilities allow users to contextualize and connect data across the platform.


Governance and Access Control

Governance ensures that data and system access are managed securely and consistently.

It includes:

  • User roles and permissions

  • Identity provider integration

  • Policy enforcement and compliance controls

This concept is applied across all layers of the platform.


End-to-End Workflow

Fynite connects all layers into a continuous workflow:

  1. Data is integrated from external systems

  2. Data is structured using models and relationships

  3. Users interact with data through configurable interfaces

  4. Activity is monitored through dashboards and visualizations

  5. AI agents analyze and generate insights

  6. Remediation agents execute actions

This unified flow enables organizations to move from data ingestion to automated execution within a single platform.


Next Steps

  • Continue to the Quick Start Guide to begin using Fynite

  • Explore Data Management to understand how data is structured

  • Learn how to configure AI Agents for automation

Last updated

Was this helpful?