Practice | Cariara Docs
Skip to content

DocsPractice

Practice

Timed practice sessions — three modes (Quick Fire, Deep Dive, Mock), per-company filtering, AI grading on a 5-dimension rubric.

Updated MarkdownOpen Practice

Overview

Practice complementsPrepare — instead of reading topics, you solve problems under time pressure./prep/practice is the timed mock-interview surface; for free-form coding without a timer, use the Coding tab inside Live.

Three timed modes

ModeQuestionsTimeBest for
Quick Fire55 minPattern recognition / warm-up.
Deep Dive315 minWorking through harder problems with breathing room.
Mock845 minFull mock-interview simulation.

Pick a mode + difficulty (easy / medium / hard) + category, then start. The timer is deliberate — under-pressure performance is what you're training for.

Problem banks

Practice draws from three problem banks:

  • Coding — 20 algorithm / data-structure problems covering arrays, strings, hash maps, two pointers, sliding window, binary search, trees, graphs, DP, backtracking, greedy.
  • System design — 12 design exercises with a guided rubric: clarify requirements, estimate scale, draw the high-level architecture, deep-dive on the hardest piece, discuss trade-offs.
  • Behavioral STAR — 12 behavioral prompts covering common rubric dimensions (leadership, conflict, ambiguity, ownership). Write the story; AI grades it.

System design whiteboard

System-design problems include an Excalidraw-based whiteboard so you can sketch your architecture as you talk through it — closer to a real on-site interview than a text-only answer.

How grading works

Cariara grades each session with AI rather than test-case execution. The grader scores your response on a 5-dimension rubric:

  • Problem solving — did you decompose the problem clearly?
  • System design — for design problems, is the architecture sound?
  • Data structures — did you pick the right tools?
  • Communication — is the explanation legible to a grader?
  • Time management — did you finish within the timer?

Each dimension scores 0–5; a session score plus per-dimension feedback comes back in seconds.

Note
Practice grading uses your free / paid AI feature allowance — coding and design attempts share the same per-feature budget as the Live helpers. SeeTop-ups & AI time for budget mechanics.

Readiness score & stats

Cariara rolls each completed session into a readiness score — a single number that tracks how prepared you are based on recent grades, dimension trends, and session count. Stats are persisted locally per device so you can track progress over time without an account dependency.

Blind 75 and Top 100

The Blind 75 and the Top 100 live in thePractice tab of SolveMe — grouped by pattern, each with an editor in six languages, tests, solutions and a guide to its pattern. Use it for systematic pattern coverage rather than timed mocks.

Need help?