AI-powered software intelligence platform Adronite raises $5M – News

Adronite Inc., the developer of an artificial intelligence codebase intelligence platform for software engineering, today announced it has raised $5 million in Series A funding led by Gatemore Capital Management to scale its product.

At its core, Adronite promises the capability to ingest codebases of any size into its context-understanding engine to provide intelligence and insights. The company says, unlike conventional tools that operate on the level of individual files, snippets or isolated vulnerabilities, it works at project-level and can reason across millions of lines of code.

This is important as the company noted that it is not constrained by context windows, which are traditionally a pain point when working with extremely large codebases. Although, many LLMs are beginning to open up larger context windows towards a scale of a million tokens, which can cover around 50,000 lines of code on average (depending on density).

“Software systems at large organizations are expanding rapidly in both scale and complexity, making a full-system view difficult to attain — let alone to implement changes safely and effectively,” said Edward Rothschild, chief executive and co-founder.

The platform itself is large language model-agnostic; developers can bring their own LLM or deployment and start from scratch or bring their own codebase. This means a developer can prompt the system to build an app or pore over an already existing project and immediately get started.

Adronite combines a number of different tools, including comprehensive reports on code quality and security. Users can receive code mapping of dependencies, data flow analysis and security frameworks to better understand how the code works together, identify vulnerabilities and remedy them. The solution also integrates with popular continuous integration and deployment pipelines.

For users used to vibe coding or Copilot-style operations, it features a method for exploration. Users can open up a dialogue with AI chat, allowing them to discover system-wide insights just by talking to an internal AI that can see the entire codebase and “talk” to their code. This can be done within an independent code editor, where the coding is done, or for exploration and insights inside the platform.

The company said the platform has near universal support for popular programming languages, covering over 20 languages including C/C++, Python, JavaScript, Rust and more.

Adronite also stressed that its focus on security goes deeper than its discovery and mapping. It also provides privacy-first deployment with the ability to bring it into private clouds or launch fully on-premises to operate locally without fear of data leaks.

“Adronite’s codebase-level intelligence, combined with its security-first deployment model, positions the company to become foundational infrastructure for some of the world’s largest and most complex software environments,” said Liad Meidar, chair of the board of Adronite.

The company said the current investment will support development for the product and scaling deployments for regulated and complex enterprise environments.

Image: News/Microsoft Designer