MacroXoft
Tech Stack

Genkit: Your Toolkit for Generative AI

Genkit is an open-source framework from Google, designed to streamline the process of building, testing, and monitoring AI-powered applications. In MacroXoft, Genkit is the backbone of our agentic systems, orchestrating interactions with generative models and external tools.

Core Concepts of Genkit

Genkit provides a structured way to build complex AI workflows. Its main building blocks are:

  • Flows: A flow is a function that encapsulates a piece of AI-driven logic. It can be as simple as a single prompt to a model or as complex as a multi-step chain involving several models and tools. All agentic actions in MacroXoft are managed as Genkit flows.
  • Prompts: Genkit allows you to define and manage your prompts in a structured way, separating them from your application code. This makes it easy to version, test, and update your prompts.
  • Tools: A tool is a function that an AI model can choose to call to get external information or perform an action. For example, our file-writing functionality is a tool that the agent can use when it needs to create or modify code in languages like JavaScript or Python.
  • Tracing & Monitoring: Genkit includes a developer UI that provides detailed traces of your flows. This allows you to see exactly what prompts were sent, what tools were called, and what the model's output was at every step, making debugging complex AI behavior much easier.

Why Genkit for MacroXoft?

Genkit is the perfect fit for our platform. Its Go/Node.js-based framework is built for production and developer velocity, and its TypeScript support integrates seamlessly with our Next.js stack. By using Genkit, we can build robust, observable, and maintainable AI agents that are easy to extend and improve over time.

AI Agentic Framework

Genkit is the engine that orchestrates the flows and tool usage within our AI agentic framework.

Explore the Framework
Generative AI Models

Genkit provides a consistent interface for interacting with various generative models.

Learn About Models