Driver app for warehouse deliveries
Why? Kurye Cebi is live with 187 votes; it matches the React Native and offline requirements directly.
Only you and the candidate see this score
iZemin brings young people, startups and organizations onto the same ground. Share your project, let the community vote, and let organizations find you through your track record.
Try it: drag the card left or right.
Driver app for warehouse deliveries
Why? Kurye Cebi is live with 187 votes; it matches the React Native and offline requirements directly.
Only you and the candidate see this score
Redesign of municipal appointment screens
Why? Sıra Bende simplifies a public appointment flow and ships its accessibility test results too.
Only you and the candidate see this score
A bot that prepares expense reports
Why? Fişle turns receipt photos into an expense sheet; a small version of the OCR and reporting work needed.
Only you and the candidate see this score
REST API for stock and orders
Why? Raf Takip has a similar stock model; prefers hybrid work while the posting is on-site.
Only you and the candidate see this score
Real suggestions are recalculated whenever profiles and needs change.
4 suggestions left
You are interested in:Nobody yet
What counts is not whether you wrote code: did you finish it, ship it, did people use it?
Share a finished project with its live link, code and "build with" tags. Built it with AI? Say so with a tag.
The community tries it, votes and leaves feedback. The week's top three on Launchpad reach the podium.
Highly rated projects become a permanent trace on your profile. Anything from your CV stays apart, marked as a claim.
Organizations publish their needs. The fit score works both ways; organizations see candidates ranked by it, and you see the postings that fit you.
Live projects get voted on and the week's top three earn a permanent badge. Votes are weighted by the voter's activity; suspicious vote clusters get flagged.
Vote and watch the ranking change.
Daily routes and proof of delivery for couriers, works offline too.
An accessible interface that cuts municipal appointments down to three steps.
A small bot that turns receipt photos into an expense sheet.
Pairs students taking the same course to study together at the library.
A stock and order API for small shops.
Projects, votes, feedback, applications and finished work add up to a score from 0 to 100, with the last 30 days weighing more. The higher it is, the likelier you are among the featured makers, and the more your Launchpad votes count.
Click the tasks and watch your score and level change.
Apprentice below 40, Journeyman below 70, Master above.
It does not change the fit score; that only says how well the two sides fit each other.
Today's tasks
Organizations often struggle to describe their problem in a structured way. iZemin takes free text, asks about missing details in 3–4 questions and produces a schema-validated need card. The budget is required and public.
We need an app so our drivers can track warehouse deliveries from their phones.
Published. Everyone who passes the filters is ranked by fit score; you pick the best 8, the top 100 or a score range.
The score works both ways: the candidate's fit to the posting and the posting's fit to the candidate's place and work-mode preferences. There is no threshold; every candidate who passes the hard filters is in the ranked list, and the organization picks the best 8, the top 100 or a score range. Only that person and that organization see the score.
The two sides meet in a harmonic mean: one weak side pulls the score down, and if either is 0 the score is 0. Activity and "Interested" do not change it.
Every step shows on a tracking page both sides share. Nobody is left wondering what happened. Click a step.
The organization structures its need with AI and publishes it with a budget. If the work fits, the person applies and can withdraw at any time.
When both sides say "Interested", it is a match. The system schedules the pre-interview within the day and time range the person chose when signing up.
A Turkish voice agent calls the person for a short introduction and sends a structured summary to the organization. If unreachable it retries three times, 30 minutes apart, then asks the same questions in writing.
Hi Baran, I'm iZemin's voice assistant. Do you have two minutes for Kıyı Lojistik's posting?
Sure, go ahead.
The app has to work offline. Have you done that before?
In Kurye Cebi; it syncs in order once the connection is back.
How many hours a week can you give?
Twenty hours, weekday evenings.
A collaboration page opens: started, in progress, done. Both sides post updates, visible only to the two of them.
When the work is done, both sides rate each other. The collaboration becomes a permanent trace on both profiles and file access closes on its own.
Sharing projects is not the only way. There are other ways to show your skills and stay part of the network.
Java, JavaScript, SQL and Python tests, plus a coding test that runs real code. Tests you pass become verified skills.
34 questions from A1 to C2 in 35 minutes. Your level becomes a profile badge; organizations can require a level.
The only assistant allowed during a test. Using AI is not forbidden; how you use it is visible to the organization.
Guess the stack, bug of the day, word and regex games. One try a day, a weekly table, and it adds to your streak.
School clubs and communities announce events, issue tickets and check people in at the door.
A teammate, a project partner, a mentor who has worked in your field or a cowork buddy; mutual matching by interests and location.
Sign up with GitHub; repos you open arrive as pre-filled suggestions to add to Launchpad.
Organizations see every person they matched with as a folder card and the shared files in one table. Sharing opens with the match and closes when the work ends.
Work mode, cities and the districts or sides of town you won't commute to. Postings that don't fit stay out of your way; the list is never shown to organizations.
There to do a job, not to decorate. Every structured output is schema-validated; anything that fails never reaches a user.
| Where | What it does | Why |
|---|---|---|
| Need structuring | Asks about missing details in free text and builds the need card. | Organizations struggle to describe their problem in a structured way. |
| Fit score | Profiles, projects and needs are embedded; similarity is one part of the candidate's fit to the posting. | Keyword matching misses synonyms and context. |
| Match reason | Writes a one-sentence "why this fits". | Makes the score understandable. Generated once, for the top 5 suggestions only. |
| Profile draft | Extracts title, summary, skills and projects from the CV and GitHub. | Sign-up takes seconds; the output stays a claim, not a verification. |
| Profile summary | Writes a two-sentence summary from projects, votes and collaborations. | The summary comes from earned traces, not from claims. |
| Voice pre-interview | Calls the person after a mutual match and sends a structured summary. | Introductions scale. If the call fails, the same questions are asked in writing. |
| Vote integrity | Flags clusters of votes on one project from freshly opened accounts with no other trace. | Launchpad rankings cannot be bought. |
Always free for young talent: profile, Launchpad, applying, pre-interviews and collaborations are never paid. Organizations start free too.
Yes. AI-assisted work is not hidden; you add tools like Cursor, Claude or v0 as "build with" tags. What counts is that the project is finished, live and used.
Only you and the organization that owns the posting. Nobody else can see it.
No. The call happens with explicit consent and no recording is stored; only a structured summary goes to the organization. If the call fails, the same questions are asked in writing.
Your general files open only in the workspace of organizations you mutually matched with, and access closes on its own when the collaboration ends. You choose organization-specific files separately.
No. Anything from your CV and links shows on your profile as a claim, kept apart from your traces. Real traces come from projects shared and voted on the platform, tests you pass and work you finish.
An organization account shows an "unverified" label on its profile until the admin team verifies it.