HyperCV: The AI-Powered CV That Actually Reads the Job Spec
Every developer I know has the same problem. They have a good CV. They spend 40 minutes manually rewriting it for each job. Then they send it. Then they hear nothing.
The CV wasn't bad. It just wasn't that job's CV.
The Insight
A CV is not a document. It's an argument. An argument that you are the specific person this specific company is looking for. And that argument has to be constructed differently for every audience.
The skills are the same. The experience is real. But the framing, the order, the emphasis — that's the work. And that work is 90% pattern-matching that humans are slow at and language models are fast at.
That's the gap HyperCV closes.
The Pipeline
HyperCV runs a three-pass process against your master CV and a job spec you paste in.
Pass 1 — Tailor
The AI selects and reframes content from your master CV to match the job. It doesn't invent anything. Every line has to exist in your history. It just surfaces the right things in the right order.
Pass 2 — Critic
This is the part most AI CV tools skip. After tailoring, the critic scores the output across five dimensions:
| Dimension | What it checks |
|---|---|
| Keyword Match | Exact terminology overlap with the job spec |
| Relevance | Most relevant experience is front and centre |
| Impact | Achievements are quantified, contributions are clear |
| Tone Match | Language mirrors the company's culture signals |
| ATS Readiness | Clean format, parseable for applicant tracking systems |
Each score is 1–10. The critic also surfaces gaps ("missing Kubernetes experience — consider addressing") and strengths ("strong open-source signal for this type of role").
Pass 3 — Refine
The refiner takes the critic's feedback and improves the CV — amplifying strengths, addressing gaps where your history supports it. Then you run the critic again. You iterate until the scores are where you want them.
This loop — tailor → critic → refine → critic → refine — is the thing. Not a one-shot generation but a feedback cycle.
The Master CV
Everything runs from master-cv.md. One file. Your full career history in clean Markdown: every role, every project, every technology, every impact metric you can remember.
The tailor pass never invents. It only selects and reframes. This means you can trust the output — if it says it, you said it first.
The discipline is keeping master-cv.md current. Every time you ship something new, add it. New role, new project, something worth mentioning in a job context — it goes in. Then HyperCV can surface it when the right job spec comes along.
Why Projects Win Interviews
The V2 roadmap adds something more interesting: outreach mode.
The insight is that traditional job applications are commodity. Thousands of people applying, optimised CV or not, all showing up in the same ATS queue. What breaks through is project-led outreach — a message that leads with something you built and connects it directly to what the company is doing.
"I'm building web3.capital — a capital allocation tool for DeFi that uses Porto passkeys to remove the seed phrase entirely. I saw you're working on the same wallet UX problem. Here's how that maps to what you need..."
That message is different because it's not a pitch. It's a demonstration. The project is the proof. The CV is the backup.
HyperCV V2 will generate these outreach messages alongside the tailored CV — using your real projects (including web3.capital, Revela, HyperDrift) as the hooks.
The Stack
- Next.js 14 App Router + TypeScript + Tailwind
- Claude CLI via
claude -p— no API key, uses your Anthropic subscription - SQLite for saved CVs with versioning and tags
- SSE streaming — all three passes stream token-by-token, so you're never staring at a spinner
The app runs locally. Your career data never leaves your machine.
What This Is Really About
Most job application advice is noise. Formatting tips. Buzzword lists. "Tailor your CV!" — without telling you how.
HyperCV is the how. It doesn't replace judgment; it accelerates it. The critic tells you what's weak. The refiner fixes it. You decide when it's good enough.
The goal is never to automate job applications. The goal is to make the signal-to-noise ratio so high that the people reading your CV can't ignore it.
HyperCV is part of the HyperDrift ecosystem — tools built to give independent developers leverage. Want to try it or talk about the outreach mode roadmap? Get in touch.
Get weekly intel — courtesy of intel.hyperdrift.io