Zencoderan artificial intelligence startup focused on software development, today announced the launch of a suite of products developed using generative AI agents designed to help software developers work faster and to create better applications.

Founded in 2023 by Managing Director Andrew Filev Zencoder offers AI agents that integrate into development workflows to provide intelligent code generation, code repair, test generation, optimization, and documentation.

Filev told SiliconANGLE in an interview that Zencoder differentiates itself from other AI-based coding assistants on the market by offering AI agents to developers. The AI ​​agents are capable of multi-step coding and self-healing capabilities, which sets the company apart from the competition.

“Tit uses original coding wizard by many developers is compared at StackOverflow on steroids,» said Filev. He added that For the most part Wizards provide code snippet suggestions and knowledge. These use cases, while extremely useful, left people feeling oversold.

To rely on AI agents as companions, Filev said his company has invested heavily in a product called Repo Grokkinga tool that allows the the company agents to analyze in depth a the company code base to understand the structure, understand the structure, identify patterns and logic behind the code. It allows agents to adapt to a the company code base to provide more relevant suggestions and better responses to queries and prompts.

“So, you can ask him in chat to fix simple bugs for you, for example, or to implement simple new features, and he will plan his work,» said Filev. “It will attempt to run this job and try to validate this work before presenting it to you.”

This highlights the second agent workflow innovation that Filev says defines Zencoder apart, what he calls agent repair. When an AI agent produces code, it will automatically analyze and refine the generated code in an attempt to fix potential bugs that could have been product or further improve its performance. It can do this in the context of the code base using its knowledge of how the code interacts with other parts of the application and will even run tests to TO DO certain it will compile without problem.

“WWe’ve invested a lot in this component that analyzes the AI-generated code and tries to fix it,” Filey said. “When it does, it uses both AI and non-AI techniques, which can be called stitching. A large part of our team has a background both in AI and things like compilers and code analyzers, so we understand code Really well and tooling how you work with code.”

When an AI agent generates codehe explainedit can use parsing to quickly discover errors and run it again with AI to find and correct errors before he is offered back to the user. Repair is a multi-step process designed to help alleviate potential hallucinations or errors it could happen during the AI ​​code generation process.

Zencoder supports major programming languages including Java, JavaScript, TypeScript, Python and C#. It can also integrate with popular editors like Visual Studio Code and JetBrains.

Filev said the company plans to launch custom agents in the future, allowing users to train an agent to automate repetitive tasks. it happens during the development process. For example, a development team may want to migrate from one version of a library to another, or from one framework to another, making many small changes throughout the codebase.

It would be Normally be done with a series of different macros then some manual modifications for everything that was lack. However, with Repo Grokking and an AI agent, much of this work could be completed asking what changes were needed to be done and the best scripting method do it.

“IIt is not an agent who claims that It is I will do all the work for you,» said Filev. “It is your companion, type of your little AI intern, who can help you recreate what you want to do.”

Image: Zencoder