There are a lot of career tools that describe themselves in sweeping terms and leave you guessing about what the product actually does. This post is the opposite of that. It is a walkthrough of SkillNode version 1 — what ships today, how the pieces fit together, and where the edges are.
If you read this and something feels unclear or missing, that is useful to us. Tell us.
Your profile starts with your resume
When you sign up, you can upload the resume you already have as a PDF. We parse it into a structured profile: your experience, projects, education, and skills, each as editable fields rather than lines trapped in a document. You can also skip the upload and build the profile by hand.
That structured profile becomes the source of truth for everything else in the app. When we tailor a resume or score a match, we are working from what you have actually done — not from whatever formatting survived a PDF export.
Tailoring for a specific job
Paste in a job description and SkillNode rewrites your resume bullets against that specific role, using your real experience as the raw material. It drafts a matching cover letter too. Nothing is invented — the point is to surface the most relevant parts of what you have done, phrased for the role in front of you.
You see the result as a live resume preview you can edit inline, with a handful of template styles to choose from. When you are happy with it, export it as a PDF. If the AI suggests a rewrite you do not like, you can keep your own version, bullet by bullet.
Matching on skills, not keywords
Most screening software does string matching: if the job says "React" and your resume says "React", you pass that filter. That breaks down fast — it misses people who wrote "built the frontend in Next.js" and rewards people who paste the job description into their skills section.
SkillNode groups related skills into clusters — more than a hundred of them — and scores overlap in tiers: a direct match, a related-skill match, or no match. When you browse opportunities, the match percentage you see reflects that capability overlap, not how many exact words you and the job posting happen to share.
Keeping the search organized
The rest of version 1 is about making the search itself less chaotic. Applications live in a tracker so you know what you applied to and when. Documents you generate are saved to a library automatically, so the resume you tailored three weeks ago is still there when the recruiter finally replies.
There is also a contact directory covering several hundred well-known employers, with links out to the right people on LinkedIn — for the part of the search that software cannot do for you. You can sign in with LinkedIn, Google, or email.
For employers
The employer side of version 1 is deliberately simpler: a dashboard for creating roles and reviewing candidates, matched on the same skill clusters students are scored against. If the matching logic is honest in one direction, it should be honest in both.
What version 1 doesn't do yet
Plenty. Interview preparation is still in progress. The employer side is thinner than the student side. Matching gets better as we refine the skill clusters, and there are rough edges we find every week — usually because a student using the beta points one out.
We would rather ship an honest version 1 and improve it in public than promise a finished product that does not exist. If you try SkillNode and something breaks or feels wrong, that feedback shapes what we build next.
Read next
How SkillNode started
Why three students in Toronto decided the resume screen was broken enough to build something better.