Rollingin
Rollingin turns an Android tablet into an offline face recognition attendance kiosk. Enroll employees, match faces on the device, review attendance in the app, and export reports for your own analysis. Everyday attendance needs no cloud service.
Designed for
Workplaces that need a shared attendance kiosk, including shops, restaurants, and offices.
The workflow
Add employee details and capture a guided, multi-angle face profile. Face quality checks run locally, and the profile and small face crop are stored inside the Android app.
Employees scan at the tablet and confirm their check-in or check-out. Face matching and attendance recording work without an internet connection, including in airplane mode.
View today’s attendance and open built-in reports on the same device. Save or share a report, or let a Super Admin export a richer data package for custom analysis.
Capabilities
Enroll employees, record attendance, browse records, and generate reports without internet access. Records are available on the device, with no cloud account needed for daily operation.
Multi-angle enrollment, face quality checks, and voice-guided scanning support the kiosk workflow.
Manage employee schedules, weekly offs, personal leave, and public holidays. Separate Super Admin and Manager roles control access; manual attendance is available when enabled in settings.
See who has checked in or out today. Super Admins can browse attendance history by date range, employee, and status, while both roles can open and save built-in reports.
Export individual reports as CSV, or use the Super Admin analysis ZIP for linked attendance data, shift summaries, and field definitions. Bring the files into spreadsheets or your own analytics workflow.
Manual Cloud Sync and cloud backups exclude face data. Full encrypted backups can include biometric records and be exported by an administrator.
Privacy & offline use
Enrollment, recognition, attendance, and reporting run locally. Daily operation sends no face or attendance data to a cloud service and needs no connection to a recognition server.
Small face crops and the face profiles used for matching live in the Android app’s private storage. Full camera frames and enrollment videos are not saved. Android automatic cloud backup is disabled.
Read the full privacy policy ↗.faceid file that includes face
crops and biometric records, and can copy it off the device.
Optional sync, cloud backup, and email password recovery need connectivity. Local attendance and reporting continue offline.
Removing an employee from active enrollment retains historical records and inactive face profiles. See the data removal guide for reset and deletion options, including exported copies.
View it in the app
Open the home screen for today’s attendance, or choose one of seven reports in Management. Managers and Super Admins can select a date range, apply filters, review summaries and tables, and reopen saved reports without leaving the app. Employee filters are available in the daily register, payroll, punctuality, and hours reports.
Missing historical schedules or attendance rules are shown as unassessed where needed. Incomplete shifts are flagged for review, so missing data does not silently become a confident attendance total.
Export & explore
Use a report as it is, build a custom dashboard, or explore attendance patterns with Claude or ChatGPT. You choose which files to export and where to use them. The app does not automatically send data to an AI service.
For everyday reporting · Managers & Super Admins
Choose a report and date range, apply the available filters, and review the result in the app. Save it for later or share a CSV through email, WhatsApp, or the Android share sheet.
Open the file in a spreadsheet to sort, chart, or pass attendance totals into your existing reporting process.
For deeper analysis · Super Admin only
In Reports → Analysis data, select up to 366 attendance dates and save a ZIP to the device. It includes linked CSVs for employees, attendance events, shifts, daily summaries, schedules, and attendance rules.
Stable field names, a field dictionary, an export manifest, and data-quality findings help custom analytics tools interpret the records and connect related tables.
For analysis with Claude, ChatGPT, or a custom tool, extract the ZIP and use the relevant CSVs with the included field dictionary and validation findings. For example: “Summarize weekly hours and late arrivals. List incomplete shifts and unassessed days separately.”
The analysis export excludes face crops, face embeddings, and credentials. It still contains employee identities and attendance records, and is an unencrypted analysis file. Review or remove identifying fields before sharing with an external service; files you upload are handled under that service’s policies.
Getting started
Designed for an Android tablet with a camera, used as a shared workplace kiosk. Get in touch to discuss device suitability and a demonstration.
Enquire about Rollingin ↗For the technically curious
React and TypeScript provide interfaces in English, Hindi, and 23 other languages through Capacitor. CameraX, MediaPipe, and LiteRT handle native capture, face quality checks, and on-device matching. SQLite stores records locally; it is protected by the Android sandbox rather than database encryption.
Read the engineering case study ↗Let’s talk
I also work with teams on custom AI products, integration, and production engineering.