Filipa is designed for conducting interviews with job candidates. The application provides
- Dual-view interface: Separate synchronized views for the interviewer and the candidate
- Question management: Load and manage interview question sets with expected answers
- Real-time assessment: Take notes and rate candidate responses during the interview
- Markdown-based: Questions and answers stored in portable markdown or JSON files
- Offline-capable: Works without an internet connection, just open
index.html - Import/Export: Questions, candidates, or individual sessions can be shared with colleagues
I was looking for a simple way to present questions to candidates during an interview. I am a big fan of the markdown format because it is easily readable in raw form and can also be simply converted to HTML with custom styles.
The main idea was to keep questions, candidate answers, and notes in a single file without needing a server or other complex infrastructure, which would be problematic to run in highly controlled environments.
I ended up with the following workflow:
- Questions can be created inside the application.
- Questions can be imported from JSON or markdown files.
- A
config.jsonfile placed next toindex.htmldefines the mapping for importing from.mdfiles. - Duplicate question detection is included.
- Sets are groups of questions used for faster assignment to Interview Session.
- Multiple candidates with multiple sessions are supported.
- Each session can represent a different question area. For example, one interviewer can cover backend questions and another can cover frontend.
- Each session keeps a copy of the questions selected from the catalog.
- Candidate sessions can be exported or imported, allowing data from two separate Filipa instances to be combined.
- Questions can be added to the current session on the fly.
- Hash-based routing allows the app to be opened directly in a browser from a local file without running a server.
- IndexedDB is used for sufficient storage space.
- A backup feature keeps all data in one place.
- The Candidate View can be opened in a separate browser window and questions can be shown from your Session page as needed.
I had this kind of application in mind for a long time, but never had enough spare time to realize it. Thanks to AI, I was able to create a first functional version I was happy with, in just 8 hours. You can also read my developer notes for this project.
I use React at work, but I chose Svelte here to keep the built application small and to step outside the React world for a while. But in reality, every single function in this project was written by AI, not me.
Try it online: https://m0jimo.github.io/filipa/
Or download filipa.zip, extract it, and open
index.html in your browser.
Filipa comes from a Czech phrase "mit filipa" which is an idiom used to describe somebody who is smart, sharp or with good practical sense. Filip/Filipa is also first name for male/female.
- Node.js (v22 or newer recommended)
- npm, pnpm, or yarn
npm installnpm run devThis will start the development server at http://localhost:5173.
npm run buildThis creates a dist/ folder with all compiled files.
After running npm run build, you can:
- Navigate to the
dist/folder - Double-click
index.htmlor open it via File > Open in your browser - The app works completely offline using the
file://protocol
- Upload the entire
dist/folder to any static hosting service - Compatible with GitHub Pages, Netlify, Vercel, AWS S3, and others
- Zip the
dist/folder - Send it to users who can extract and open
index.htmllocally - Ideal for regulated environments where internet access is restricted
MIT (c) Ota Vaclavik — see LICENSE for details.
Dedicated to Filipa, after whom this application is named.
