AI-Assisted Software Builder · Open-Source Contributor

I've been building software since 2013, when I wrote my first line of C.

I took an unconventional route into software. I spent years learning, experimenting, and building projects, but never established a traditional software engineering career or turned those projects into successful businesses.

AI changed what I could execute.

Today, I use AI extensively to investigate codebases, solve engineering problems, evaluate implementations, and contribute to software used by other developers. I'm putting that approach to the test in established open-source projects, where changes have to work within existing architectures and withstand tests, CI, and maintainer review.

These are contributions to existing production codebases, not projects I created independently.

I use a multi-model AI workflow involving Claude, GPT, Gemini, Copilot, as well as open source models.

Rather than relying on one model's first answer, I use multiple systems to investigate problems, propose approaches, challenge assumptions, and review potential solutions.

My process typically involves:

• Understanding the issue and investigating the relevant codebase.
• Reproducing the reported behavior and identifying its cause.
• Comparing possible implementations and examining trade-offs.
• Developing changes and regression tests.
• Running available tests and validation checks.
• Incorporating feedback and iterating through review.

AI is an explicit part of my workflow. I disclose its use in the pull requests I submit.

The tools help me work. I remain responsible for understanding, evaluating, and standing behind the changes I propose.

I'm working across open-source projects in developer infrastructure, AI tooling, and application frameworks.

View all my open-source pull requests →

Between late 2025 and mid-2026, I built four AI products. None became commercially successful, but each gave me practical experience taking software from an idea to a working implementation.

• HireProof — Resume optimization with honesty checks and an LLM evaluation suite.
• PlainTextOps — AI email workers for Google Workspace.
• Sales Layer — Sales call analysis grounded in transcript evidence.
• Epistemic — A concept for adversarial, multi-model verification of AI code review.

I believe AI is changing how software gets built, and that the ability to direct, evaluate, and verify AI-generated work will become an increasingly important engineering skill.

I'm building that capability through real contributions, public feedback, and increasingly challenging problems.

I'm looking to work with an engineering team in a full-time, part-time, or contract capacity.

I don't claim a conventional background. I invite people to look at the work.

GitHub: github.com/jadibdev