SwapADay

AI is reshaping careers. Remix your career like a playlist. Blend skills, tastes and networks with friends and role models to co-create projects, swap real-world experiences and explore what's next.

Comment

Inspiration

SwapADay started with a simple belief: people find it hard to change careers not because they lack information. They lack exposure. Reading about a job tells you very little. Spending an hour inside someone's real work tells you a lot.

Earlier this year I ran a pilot where mid-career professionals tried real work through short "swaps" with hosts: watch, do a real task together, or ask what the job is really like. [X hosts, X swaps, one line on what you learned.] People loved it. But every host and every user came through me, and a swap was a one-off with nothing after it.

Then I noticed how career changes actually happen: often through a project with someone you already know. That's where the new direction came from. What if your career wasn't just your skills and job title, but also what you're curious about, who you know, and what you could build with them?

What it does

SwapADay Blend is like Spotify Blend for your career, with Hinge's sense of intention.

  • Your profile holds what you've proven (with the evidence behind it), what you're curious to try, the fields you care about, and your network organized into playlists around goals ("Raise a pre-seed," "Pilot a program with a college").
  • Blend puts you next to a friend or someone you follow. You see where you overlap, what each of you brings, and projects you could make together. You also see a blended playlist of who you'd call from both networks, and the intro you didn't see: two people, one from each network, who have never met and together cover what the project needs.
  • Swap lets you test it before committing: a short session inside the other person's real work. Swaps are anchored to a real project, so each one has a reason and a next step.
  • Explore lets you browse fields, people, swaps and projects before you ever sign up.

Blending networks with a friend is automatic. With someone you follow, you ask for a specific idea, and they can say no.

How we built it

Before today, I had the pilot, a swap room and a host pipeline. Today I built the discovery layer on top.

I vibe-coded the prototype with Claude as a single-page web app: onboarding, explore, blend, swaps, projects and playlists, with generated cover art for each field. To test the "bring your LinkedIn" path, I ran it on my own data: my profile became skills with evidence, and my 2,473 connections were sorted by job title into fields and turned into playlists.

Challenges we ran into

  • Balancing two models. Blend is about discovery, and Swap is about commitment. It took several rounds to land on this: explore freely, blend to see fit, swap to commit.
  • Onboarding without homework. Early versions asked for a profile or a full network import before showing anything. The fix was letting people explore first, then asking for one thing at a time, only when it pays off right away.
  • Privacy. Blending networks means handling people who never signed up. Network blending needs consent for each idea, and people's own playlists stay private by default.

Accomplishments that we're proud of

  • The loop holds together: explore, then blend, then swap, then blend smarter.
  • On my own network, the app found introductions I hadn't thought of.
  • It turns a founder-dependent marketplace into something that can spread. Every blend brings in a second person, and "who's missing" pulls in a third.

What we learned

  • Blend before you swap. A swap with a reason and a next step is far more compelling than a one-off.
  • Ask for one, not all. Playlists should grow from moments ("need a teacher? add one"), not from a big import at the start.
  • Show first, infer second. Skills should come from what people did, and swaps are how a skill gets seen and confirmed.
  • Consent has to be specific. "Blend networks with me for this idea" gets a yes more easily than "share your contacts."

What's next for SwapADay

  • Pilot with [community] and measure three things: how many blends become swap requests, how many swaps happen without me setting them up, and how many people invite someone new.
  • Connect blends to the existing swap room, so a proposed swap becomes a real session.
  • Build proper project idea generation and matching, the hardest technical piece, with our engineering lead.
  • Bring AI companies in as project hosts, so professionals get hands-on AI exposure in their own field and companies learn that field's real pain points.

Built With

  • claude
  • netlify
Share this project:

Updates

Submission history