Measure. Map. Close. Prove.
Gapmap is built around one idea: you can't fix gaps you can't see, and you can't trust gaps you self-assessed. Here is the loop, end to end.
Test: a 15-minute adaptive diagnostic
Pick your framework and go. No signup, no email. The diagnostic is an adaptive ladder: every topic starts at middle difficulty, a correct answer sends you up toward senior, a miss sends you down. Two questions per topic is enough to place you on a band, because the questions are written around real misconceptions, not trivia.
- Anonymous: your session lives in a cookie until you decide to save it
- Deterministic and instant: answers are checked by code, so results appear in seconds
- Distractors encode real misconceptions, so a wrong answer tells us which one you hold
Map: your level, per topic, on one picture
The result is a map, not a score. Areas radiate from the center, topics orbit areas, subtopics orbit topics. Every measured topic gets an honest band from junior to senior; everything unmeasured stays gray. The gray is the point: it shows the shape of what you haven't proven yet.
- Bands per topic: junior / middle / middle+ / senior
- The syllabus is derived from roadmap.sh, the interview handbooks, and real 2025-26 loop reports
- Verified topics get a ring; the rest wait for you
Learn: study content is free, all of it
Every live topic ships notes at three depths (junior, middle, senior), written for interview preparation rather than for beginners. Read them without an account, straight from the topic pages. The business model is simple: knowledge is free; measurement and coaching are what you pay for.
- Three depths per topic, so you read at the level you're closing instead of starting from zero
- Weakest-first ordering on your map tells you where to start
- Personalized plan to your interview date: coming
Prove: an AI interviewer that pushes back
A topic closes only when you pass verification: free-form answers graded against a senior rubric with weighted criteria, senior signals, and red flags. If your score lands near a band boundary, the grader does what a real interviewer does: asks a follow-up targeting your weakest point, then grades the whole exchange. No self-assessment, ever.
- Rubric-graded: explanation counts, keyword-dropping doesn't
- A fundamental misconception caps your score below the band; fluency won't save it
- Voice-to-voice mock interviews: coming as the flagship
The diagnostic is free and takes 15 minutes.
Walk out with your map and the exact list of topics that would have failed you.