Most interview platforms are puzzle grinders. O(1) Prep puts you in a real interview.
- A back-and-forth, not a quiz - the interviewer follows up, pushes back on your reasoning, and adds constraints mid-session just like a real one would
- Feedback that actually tells you something - you get written scores and specific critique on your approach, code quality, communication, and tradeoffs - not just pass/fail
- Your own account - sign in and your sessions, history, and code are saved to your account in the app's database; self-host it and the data stays on infrastructure you control
- Bring your own engine - runs on your local Claude Code CLI by default (no API key), or point it at OpenAI or OpenRouter (one key, hundreds of models)
- 150+ problems grounded in real engineering - each one has a real-world scenario, not just "given an array..."
The repository is a monorepo: the Python API lives in backend/, the web client in frontend/, and shared infrastructure (docker-compose.yml) at the root.
git clone https://github.com/shettydev/o1prep.git
cd o1prep
docker compose up -d # start PostgreSQL (from the repo root)
cd backend # the Python API lives here
python3 -m venv venv && source venv/bin/activate
pip install -r requirements.txtCopy backend/.env.example to backend/.env and set your values. By default the AI engine is your local Claude Code CLI — no API key, just run claude once to sign in. Set AI_PROVIDER to openai or openrouter to use those instead (see .env.example for their keys).
DATABASE_URL=postgresql+psycopg://o1prep:o1prep@localhost:5433/o1prep
SECRET_KEY=change-me
AI_PROVIDER=claude # claude (default) | openai | openrouter
# OPENAI_API_KEY=sk-... # only when AI_PROVIDER=openai
# OPENROUTER_API_KEY=sk-or-... # only when AI_PROVIDER=openrouter
Create the schema, load the problems, and run — all from backend/:
export FLASK_APP=app
flask db upgrade # create tables
python scripts/seed_problems.py # load the problem library
python app.pyThis serves the API (and the legacy server-rendered UI) on http://localhost:5000.
The web client is a separate Next.js app in frontend/ that talks to the Flask
API. In a second terminal:
cd frontend
pnpm install
pnpm dev # http://localhost:3000Open http://localhost:3000. The dev server proxies /api/* to Flask on
127.0.0.1:5000 (see frontend/next.config.ts), so the session cookie stays
same-origin and auth works with no CORS setup. Point it elsewhere with
API_ORIGIN=http://host:port pnpm dev.
On macOS, the AirPlay Receiver squats on port 5000 over IPv6 — the proxy targets
127.0.0.1(IPv4) specifically to avoid it. If you change the Flask port, updateAPI_ORIGIN.
Prerequisites: Python 3.11+, Node.js 20+ and pnpm, Docker (for PostgreSQL), an AI engine (the Claude Code CLI needs no key; OpenAI/OpenRouter need an API key), and a modern browser (Chrome/Firefox/Safari).
- Picking a Problem
- Study Mode
- Running an Interview
- Voice Interviews
- Code Editor & Execution
- How the Interviewer Works
- Your History & Progress
- Problem Library
- Keyboard Shortcuts
- FAQ
- Contributing
- License & Attribution
Browse and filter 150+ problems to find what you want to practice. Filters on the left let you narrow by category, difficulty, or whether you've attempted a problem before. Status dots on each card show how you've done - green for a hire signal, yellow for mixed, red for no hire.
Not sure what to pick? Hit Surprise Me to start a random interview from whatever's currently visible.
Search - Filters by title, summary, or skills. Cmd+K / Ctrl+K opens a full command palette.
Category tabs - Filter by topic: stateful, parsing, scheduling, search, streaming, infra, concurrency, api_design, syntax, arrays, strings, linked lists, trees, graphs, dynamic programming, backtracking, debugging.
Difficulty pills - Multi-select: Easy, Medium, Hard.
Warm Up toggle - Shows only shorter, lower-stakes problems.
Sort tabs - Default, Easiest first, Hardest first, Not attempted, A-Z.
- Click Study to read it in full and chat with an AI tutor before committing to an interview
- Click Practice to jump straight into a mock interview
- Click a skill tag to filter the list to related problems
Fuzzy search across all problem titles, summaries, skills, and categories. Enter to start Practice, Cmd+Enter / Ctrl+Enter to open Study Mode, Escape to close.
Read a problem fully and chat with an AI tutor before you practice - useful when a topic is unfamiliar or you want to think through approaches first.
Two panels side by side — drag the divider to resize the tutor chat, and its width is remembered. The left panel has the full problem - scenario, constraints, examples, key skills, and follow-up challenges. The right panel is a tutor chat where you can ask anything about the problem without being handed the solution.
Good things to ask the tutor:
- "What data structure would work for O(1) lookup and O(1) deletion?"
- "What's the difference between BFS and DFS here?"
- "What edge cases should I consider?"
When you're ready, go back and hit Practice to start the interview.
Three panels: the interviewer chat on the left, your code editor in the center, and an optional tutor sidebar on the right. On a narrow screen, tabs switch between the chat and the editor.
The interview runs like a real one - you'll clarify the problem, talk through your approach, then implement. The interviewer watches what you submit and responds to it. You can submit code multiple times; each submission gets reviewed.
- Write your solution in the editor
- Click Submit Code
- Your code runs against test cases automatically
- Results appear in the chat (X/Y passed, with per-test details)
- The interviewer reads the results and continues from there
Click Tutor in the top bar to open a hint sidebar. It runs separately from the interview conversation so you can ask for conceptual help without disrupting the flow. It won't give you the answer directly.
The interviewer gives a structured debrief with a hire/no-hire rating, scores across dimensions like code quality, communication, and problem framing, and specific written feedback on what was strong, what was missing, and what a better answer would have looked like.
Status: temporarily unavailable in the web client. Voice mode is being reworked to route through a natural-sounding neural TTS provider — the browser's built-in speech synthesis sounded too robotic for a realistic interview. The pipeline (browser speech-to-text → your chosen LLM → text-to-speech) is in place behind the scenes and will return once a good TTS voice is wired in.
The intent: switch to Voice mode before starting and talk through your solution out loud — your speech is transcribed live, the interviewer replies through your speakers, and the whole session is saved to your history just like a text one, all while you can still write and Submit Code.
A CodeMirror editor (Python, JavaScript, or TypeScript, picked from the language selector) with syntax highlighting, auto-closing brackets, line numbers, and smart indentation. Switching language loads that language's starter code and runs against its test suite. Your code is auto-saved as you type - you won't lose it if you navigate away.
| Button | What it does |
|---|---|
| Run | Executes your code and shows stdout/stderr |
| Run Tests | Runs test cases against your solution and shows pass/fail per case |
| Clear | Clears the editor (asks for confirmation) |
Note: Run Tests only works during an active interview - it uses the problem context to generate cases. Run Code works anytime.
The interviewer runs a structured session, not a quiz. It asks you to clarify the problem, discuss your approach before you code, and then works through the implementation with you - adding constraints, asking follow-up questions, and adjusting based on what you say.
| Phase | What happens |
|---|---|
| Opening | Problem is presented; you ask clarifying questions |
| Approach | You walk through your plan before writing any code |
| Implementation | You code; the interviewer interjects and follows up |
| Follow-ups | Harder constraints or variations if things are going well |
| Testing | You discuss edge cases and test coverage |
| Debrief | Structured feedback and a hire signal |
| Rating | What it means |
|---|---|
| Strong Hire | Exceptional performance |
| Hire | Solid; meets the bar |
| Lean Hire | Good with some gaps |
| Mixed | Some strengths, some concerns |
| Lean No Hire | Concerns outweigh strengths |
| No Hire | Did not meet the bar |
Written feedback covers: what was strongest, what would be a concern, what a better answer would have looked like, and 3 concrete areas to work on next.
Click History in the top bar to see all past sessions. Click any entry to reload it - your messages and last submitted code are fully restored. You can continue the conversation from exactly where you stopped.
Click the X / Y done chip to see your progress by category.
Each category shows a progress bar and lists the problems you've attempted with their ratings. Status dots on problem cards update as you practice:
- Empty - not attempted
- Green - Hire or Strong Hire
- Yellow - Lean Hire or Mixed
- Red - Lean No Hire or No Hire
150+ problems across 18+ categories:
| Category | Focus |
|---|---|
| Stateful | LRU caches, time-based KV stores, undo/redo buffers, versioned state |
| Parsing | Config parsers, expression evaluators, template engines, dependency graphs |
| Scheduling | Task schedulers, rate limiters, job queues, interval merging |
| Search | File crawlers, in-memory search indexes, shortest path, autocomplete |
| Streaming | Moving averages, top-K frequent items, deduplication, windowed aggregation |
| Infra | Connection pools, retry with backoff, batch coalescing, circuit breakers |
| Concurrency | Thread-safe queues, worker pools, consistent hashing, rate-limited fetchers |
| API Design | Cursor pagination, query builders, plugin registries, diff/patch engines |
| Python Syntax | Loops, list comprehensions, slicing, built-in idioms |
| Arrays | Two sum, prefix sums, sliding window, binary search |
| Strings | Reversal, anagram detection, longest common prefix, pattern matching |
| Linked Lists | Reverse, merge sorted lists, cycle detection, nth from end |
| Trees | Level-order traversal, validate BST, lowest common ancestor, max path sum |
| Graphs | Number of islands, shortest path, topological sort, cycle detection |
| Dynamic Programming | Coin change, edit distance, longest increasing subsequence, knapsack |
| Backtracking | Subsets, permutations, N-Queens, word search |
| Debugging | Find and fix bugs in broken implementations |
- A real-world engineering scenario explaining why this problem comes up
- Alternative contexts that use the same pattern
- Formal problem statement and constraints
- 2-3 worked examples with input/output
- Key skills and follow-up challenges
- Starter code, pre-written test cases, and a full solution explanation with complexity analysis
| Shortcut | Action |
|---|---|
Cmd+K / Ctrl+K |
Open command palette |
Escape |
Close palette, drawers, or modals |
↑ / ↓ (palette) |
Navigate results |
Enter (palette) |
Start Practice on selected problem |
Cmd+Enter / Ctrl+Enter (palette) |
Open Study Mode |
Enter (interview chat) |
Send message |
Shift+Enter |
Insert newline |
Tab (editor) |
Insert 4 spaces |
Do I need a paid AI account?
Not necessarily. The default engine is your local Claude Code CLI, which runs on your existing Claude sign-in with no separate API key. If you prefer, set AI_PROVIDER=openai or openrouter and supply that key instead.
How much does each interview cost? On the default Claude CLI engine it draws on your Claude usage rather than a per-call charge. On OpenAI or OpenRouter it's pay-per-token — roughly $0.10-$0.50 per text session depending on length and the model you pick.
Can I use other programming languages?
Yes. Pick Python or JavaScript from the language selector in the editor; the code
runner and test harness support both. JavaScript runs through Node.js (install
it, or set NODE_BIN). TypeScript is also wired up (via tsx/TS_CMD).
Problems are translated per language, so the selector offers whichever languages
a given problem has been authored for.
Can I pick a different model? Yes. The Settings (gear) menu lets you choose the model and reasoning effort for new interviews and tutor chats. On OpenRouter you can search its live catalog of hundreds of models; on Claude and OpenAI you pick from that provider's list.
What happens if I close the browser mid-interview? The session is auto-saved every time you send a message or submit code. Open History to resume from where you left off.
Can I retake the same problem? Yes. Multiple sessions on the same problem are tracked separately. The status dot reflects your best performance.
Why is the feedback so detailed and critical? That's what makes it useful. Vague feedback doesn't help you improve. The debrief is written the way a hiring committee would actually talk about your performance.
Where is my data stored? In the application's PostgreSQL database, scoped to your account. If you self-host, that database runs on infrastructure you control. Your messages and code are sent to your configured AI provider (Claude, OpenAI, or OpenRouter) to power the interviewer.
Can I add my own problems? Yes - see CONTRIBUTING.md for the full problem YAML format.
O(1) Prep is released under the MIT License.
This project is a fork of codingprep by Amrutha Gujjar (MIT-licensed at the time of forking), with substantial modifications. The original work's copyright notice is preserved in the LICENSE file, as required by the MIT license. The upstream project has since been made closed-source; this fork continues independently from the MIT-era snapshot and is not affiliated with or endorsed by the original author.
Questions or feedback? Reach out on LinkedIn.