.png)
In November 2024, Andrew Christianson was working on his startup fictie.ai, an AI-powered audiobook application, when he realized that existing AI coding agents, like Windsurf and Cursor, were up-charging and limiting his compute. Instead of accepting the limitations, Andrew developed RA.Aid to cut out the middle man and plug directly into any model, including high-power reasoning and open weight models. Just a few months later, RA.Aid would become the first project accepted to Catalyst, OCV’s sponsorship program that provides funding and mentorship to open source authors and maintainers to catalyze project growth over a three-month period. Now in it’s final week in the program, RA.Aid has evolved into a mature AI coding agent that combines research, planning, and implementation to help developers build software faster and smarter.
"The project wouldn’t be in its current state without the Catalyst sponsorship program, especially in this timeframe. It was an experimental side project," said Andrew. "Once I joined and was able to deliberately work on building out RA.Aid, that's when it started to grow.” Since joining the program, RA.Aid's weekly downloads jumped from 700 to 4,400—a sixfold increase. GitHub stars climbed from approximately 200 to nearly 2,000, and a thriving community has developed around the project, with 400 active users now participating in its Discord server.
Not just another coding assistant
While most AI coding tools focus primarily on generating code, RA.Aid captures multiple phases of the software development lifecycle. It takes a prompt and conducts deep research on the codebase and current documentation for the tech stack, then it plans a multi-step workflow for development, writes the code, and tests the code as it writes it.
After completing the three-month program, RA.Aid can now make changes to existing projects, can develop data analysis programs with Python, and even build full-stack applications with web development technologies like TypeScript and Next.js. It helps users create and deliver production-grade, high-value changes to their codebases more efficiently than existing autonomous coding alternatives.
Whenever you give it an assignment or a task, it will autonomously crawl your code base, do deep research, and formulate research reports. Once it creates the research reports, it starts a planning process. In addition to its core functionality, the project now has several differentiated features like a native multitasking system, terminal command integrations, and persistent memory. If you give RA.Aid a complex assignment, it will break it down into multiple tasks. Its memory has a garbage collection process, so when it accumulates enough memory, it will prune those down to the core.
“The more that you use RA.Aid against your code base, the more it learns about your code base,” said Andrew. “It’s similar to if you had a software engineer or a coder working on your code base.”
The Catalyst Effect
It was the program’s structured approach and mentorship that transformed RA.Aid from a side project into a thriving open source tool with commercial potential, according to Andrew. "The deliberate focus on setting weekly goals and working to meet those goals every single week was extremely valuable,” said Andrew. “The OCV team helped me define my goals, provided perspective, and added a sense of external accountability to the project." As the Catalyst program comes to a close for RA.Aid, the project has positioned itself as a serious contender in the AI coding agent space.
"It was a wild success and I am very happy to have been part of it," said Andrew. "Even if a company didn't come out of it, it was a wild success because an open source project grew 10x plus that wouldn't have otherwise grown nearly that much."
For other open source creators considering applying to Catalyst, RA.Aid stands as a powerful example of what focused development, mentorship, and structured support can achieve in just three months.
OCV's Catalyst program provides funding and mentorship to open source authors and maintainers to catalyze project growth over a three-month period. The program aims to support open source projects with commercial viability by helping maintainers increase project usage, revitalize project activity, grow community contributions, and develop relationships with experienced open source experts.
If you have an open source project with commercial potential, Catalyst could help you achieve similar growth. If you are an open source maintainer or author interested in the Catalyst program, you can apply for sponsorship here: Apply Now.