Introducing Machinet: AI Assistant For Developers

In the world of software development, engineers face numerous challenges daily, from writing lines of code to catching bugs and generating unit tests, all while maintaining creativity. This can be daunting.

AI assistant tools like Machinet AI emerge as a secret weapon, serving to automate the routine, overcome productivity barriers, elevate efficiency, and make the coding journey more smooth in general. In this article, we’ll explore what Machinet AI is and how developers can effectively use this tool to streamline their workflow. 

What is Machinet AI?

Machinet AI is a practical coding assistant designed to help developers increase productivity and write code with fewer errors. As a plugin, it seamlessly integrates with your coding environment and becomes an integral part of your workflow. It's not just a tool but a helpful AI companion. Here's how Machinet AI can assist you with your daily workflow:

Seamless Debugging and Code Crafting

Machinet AI is designed as an assistant that effortlessly mends code, fixes errors, and even crafts new code, fitting precisely into the fabric of your ongoing project. This is the proficiency that Machinet's context-aware AI Chat offers. It understands your project in-depth, using a sophisticated indexing mechanism to ensure that every piece of code it provides or modifies is contextually apt and coherent with your entire project.

Context-aware AI Chat

Swift and Meaningful Unit Test Generation

Machinet Unit Test AI Agent saves hours on manually crafting unit tests, simplifies the process, and utilizes context to create meaningful unit tests that align with the behavior of your code.

Using popular frameworks like JUnit and Mockito assures that the tests are not just comprehensive but also structured in a Given-When-Then style, enhancing understandability and maintainability.

Unit Test Agent

Machinet AI is designed to understand your entire project, provide assistance, and augment your coding experience, making it not only efficient but also surprisingly joyful.

Machinet AI vs ChatGPT

One familiar with the capabilities of ChatGPT can ask: Why do I need a special tool for coding assistance? Let's explore why Machinet AI stands out as the preferred choice for developers, providing a more optimized, efficient, and integrated coding assistance experience.

Reason #1. Tailor-Made for Developers

While ChatGPT casts a wide net, grasping a multitude of topics with generalist knowledge, Machinet AI sharpens its focus exclusively on developer tasks. Not just a tool for chat or content, Machinet AI provides functionalities like code generation, error resolution, and unit test creation, each intricately designed for developer-centric use cases, ensuring more accurate and relevant results.

Reason #2. Contextual Awareness

The architecture of Machinet AI is designed to do more than just understand individual requests; it maintains a rich awareness of the broader project context. This means that the code snippets, tests, and solutions it offers are not isolated responses but are deeply and meaningfully aligned with your existing codebase without the need to give additional context. 

Machinet AI is aware of the currently opened file, all references to the current file, the project structure, and if you opt in to the "Index" functionality, your entire codebase. When generating tests, it also takes into account the libraries and frameworks you're using.

In this part, it’s important to note how this fits within a size-limited LLM context. The truth is, we don't input everything into the context all at once. Instead, we let the LLM decide what it needs. This is the "Dynamic context" feature in action. You can always check what was populated into an AI context using the info panel located in the top-right corner of each message.

Let’s consider an example: assume we have a typical layered architecture and need to implement a new service method and expose it via a REST controller. Thanks to contextual awareness, you can focus solely on the requirements of the method in your message, and Machinet AI will handle the rest:  

AI chat contextual awareness

This isn't a case of LLM hallucinations or speculations. We had only the StudentService class open, yet AI correctly recognized where the Student's email was located. It also replicated the style and approach of exposing REST endpoints in the StudentController. To verify that it worked with the existing classes, let's take a look at the info panel that we mentioned earlier: 

AI context info panel

Reason #3. Seamless Integration into Your Workflow

As a plugin, Machinet AI embeds itself into your workflow, providing access to AI Chat and the Unit Test Agent without the need to leave your coding environment. The need to toggle between tabs and provide auxiliary instructions is eliminated, thereby reducing friction and keeping your focus solely on the code at hand.

You might wonder if there is really such a big difference. To understand its impact, let's consider a typical developer's task. Imagine you're tasked with implementing or modifying a method within a service. Typically, a service relies on various dependencies, such as repositories, integration clients, message producers, entities, DTOs, converters, and more. And don't forget, a single message often has a limited length, potentially making it insufficient to contain all the required information. As a result, you may find yourself having to explain to the LLM what you need meticulously.

Once you've managed to assemble all the pieces and receive a solution from ChatGPT, the work isn't over. Your solution might encompass multiple files, requiring you to shuttle back and forth between the chat and your IDE to integrate the code snippets into your project.

However, with the Machinet plugin, as previously mentioned, you don't need to prepare the context manually. When you receive a solution, all your files are conveniently accessible in the same window within your IDE.

Beyond these limitations, LLMs, including Machinet AI, sometimes make mistakes, such as suggesting non-existent APIs. However, it's important to note that the quality of the context you provide greatly influences the accuracy of LLM responses. In Machinet AI's case, as we’ve discussed above, it excels in offering accurate solutions. But let's assume you encounter errors that need resolution after applying the initial solution to your code. In this scenario, you'd have to return to the chat window to address these issues with ChatGPT again.

With the Machinet AI plugin, you have the option to discuss and address the errors within your current file swiftly, and if preferred, you can use the chat window for the same purpose. 

Analyze errors

After your code is compiled, your task is still not complete since it needs to be tested as well. One more round of copy-pasting awaits you with ChatGPT. But here's the most interesting part: Machinet AI simplifies this process with a one-button solution. Simply click the arrow on a class, and you'll shortly receive a merge window with proposed tests. 

Merge new tests

Reason #4. Experience the Power with a Free Plan

The Machinet AI journey commences with a free plan, opening the door to a streamlined coding experience. With 20 free Unit Test Agent tests and AI Chat available, you can explore and experience the tangible benefits it brings to your development process.

‍

Starting With Machinet AI

To start, simply install the Machinet plugin for JetBrains IDE and embark on a path toward a more empowered and efficient coding experience.

You can also join our Discord community, where developers, tech enthusiasts, and the Machinet AI team come together to discuss, share insights, and provide support related to the tool and software development practices. 

Simplify your coding journey with Machinet AI!

‍