Resumebut AI ✦ ✧ ✦ ✧ ✦
Chat with my résumé, or tailor it to your role
Two ways to dig into my background, both running entirely in your browser — nothing you type is ever sent to a server. Chat to ask questions about my experience, skills, and projects, or Tailor to paste a job description and get a resume reordered and rewritten for that role.
Chat with my Résumé
Ask anything about my experience, skills, or projects. Runs entirely in your browser — nothing you type is sent to a server.
Try asking:
Resume & AI Tailoring
Paste a job description below to create a tailored resume, or view the standard version below.
Under the Hood
🧠 Model
Llama 3.2 3B Instruct — Meta's small instruction-tuned model, 4-bit quantized
(q4f32_1) so it fits comfortably in browser GPU memory. The weights (~2 GB)
download once and are cached by your browser, so the first run takes a few minutes and reruns
are fast.
⚙️ Engine
WebLLM from the MLC-AI project, running on WebGPU. The model executes directly on your GPU inside the browser — no backend, no API keys, no cloud inference. Requires a WebGPU-enabled browser like Chrome or Edge 113+.
🔀 Pipeline
Relevance matching, experience ordering, and skill priority are scored deterministically against my real resume data. The LLM makes one grounded prose call for the title, summary, and fit analysis — which keeps a small model from hallucinating job requirements into my experience.