Using career data as AI context for smarter job search assistance

One of the most immediately practical applications of the personal context data approach I've been developing is career support. Think about what happens when you ask ChatGPT to write you a cover letter: you paste in the job description, maybe add a few lines about yourself, and the model generates something that's grammatically perfect and utterly generic. It doesn't know that you transitioned from traditional PR into tech marketing and that the transition story itself is a selling point. It doesn't know that you've spent the last year building AI tools and that your hands-on experience makes you different from every other marketing person claiming to "leverage AI." It doesn't know you're based in Israel, work remotely, and have specific timezone constraints. All of that context exists in your head, and without it, the AI produces cover letters that could belong to anyone. The Career Data AI Context Repo is my solution: a structured repository of career information designed to be ingested by AI assistants so they can generate genuinely personalised career materials.

How the context store is structured

The repository is a modular collection of markdown files, each containing a discrete, standalone piece of career context. There are files covering professional experience, skills and competencies, career aspirations, work preferences (remote, timezone flexibility, deal-breakers), and the kind of nuanced positioning information that makes the difference between a generic application and one that lands. My resume is available both as a structured JSON file and at my personal website, so AI tools can access it in whatever format works best for their retrieval pipeline. The whole thing opens with a greeting addressed to the AI bot that will be reading it — which I find amusingly practical — along with instructions to read the markdown files, use them in combination with user prompts, and prioritise accuracy and sensitivity to the nuances of the professional background described.

From dictated notes to structured context

The context data was initially gathered through dictated notes transcribed with AI — I literally talked about my career for twenty minutes into a voice recorder, then processed the transcript into structured context snippets. This turned out to be a far more natural and productive way to capture career context than sitting down and writing it out, because speaking triggers a different kind of memory recall than writing does. You end up mentioning projects and experiences that you'd never think to include in a written document but that are exactly the kind of colour that makes AI-generated career materials feel authentic rather than templated. The intended use cases include cover letter generation, resume version tailoring for specific roles, general career opportunity suggestions, and other career-related advisory tasks. The repo is open source on GitHub for anyone who wants to build something similar for their own career data.