What it is
The job market puts professionals in an impossible position: tailoring your resume for every application is the difference between getting noticed and getting ignored, but the process is exhausting, and the moment you hand it to an AI, your voice disappears into something generic. Helm solves this by making your authentic stories the foundation rather than the AI’s.
The core loop: paste a job description, surface the skills it requires, capture the experiences that are relevant in your own words. Those stories go into a career library that compounds over time, saved once for one application and reusable for every application after it. The outputs stay authentic because the source material always was.
The next phase takes the library further. Each story, skill, and achievement becomes a node in a personal career knowledge graph, an agent-managed record that grows richer as you add to it. New inputs can be anything: a voice memo after a tough interview, a resume upload, a description of the role you’re working toward. The agent structures each one, connects it to what’s already there, and surfaces the gaps worth paying attention to. The longer you use it, the better it knows you.
Why I built it
Most professionals enter a job search the same way: dig up last year’s resume, spend hours trying to remember what they actually did, then either spend days manually tailoring it or hand it to an AI and get back something that technically checks the boxes but sounds like a press release.
The problem isn’t effort; career knowledge tends to live in the wrong place, scattered across old documents, half-remembered in performance reviews, fully formed in the stories people tell in interviews but never written down anywhere useful.
Early user research kept surfacing the same pattern: every person who used AI to help with their resume felt compelled to edit the output because it sounded too much like ChatGPT. One user put it perfectly when she said she was afraid to “outsource her voice.”
That reframe changed the direction of the product. The question shifted from “how do we generate better bullets” to “how do we make sure the AI is working from authentic material, not generating from scratch.” Once stories exist in structured form, persistent and connected to skills and roles, they become something more interesting than a resume input: a compounding record of what someone has actually done and is capable of.
Status
The job description analysis, story capture, and resume bullet generation loop is working, with deeper coaching features currently in design.