This project automates the process of unmatching users on Bumble, removing repetitive manual steps and ensuring consistent cleanup of unwanted matches. By leveraging Android automation, Bumble Auto-Unmatch helps users maintain a tidy inbox and streamline dating-app workflows with minimal effort.
This automation tool performs hands-free unmatching actions on the Bumble Android app. It removes the tedious routine of manually navigating profiles, confirming unmatch dialogs, and clearing histories. For users or teams managing large-scale dating-app workflows, it provides fast, reliable automation that improves efficiency and consistency.
- Reduces time spent manually browsing and unmatching inactive or undesired profiles.
- Uses Android UI automation to reliably trigger navigation, menu actions, and confirmations.
- Supports scheduled or on-demand cleanup sessions.
- Operates without requiring device rooting or intrusive modification.
- Designed to scale across device farms and automation clusters.
| Feature | Description |
|---|---|
| Automated Unmatch Workflow | Navigates to match list, opens each profile, and performs unmatch actions. |
| Smart UI Detection | Uses UI Automator/Appilot signals to identify correct buttons and dialogs. |
| Retry Logic | Automatically retries unfinished actions due to slow device or network delays. |
| Session Scheduler | Allows scheduled cleanup runs at configurable intervals. |
| Device Abstraction | Supports ADB-less and Appium-like flows across multiple Android devices. |
| Logging & Telemetry | Outputs structured logs for each unmatch attempt and result. |
| Parallel Device Execution | Runs multiple devices simultaneously using horizontal workers. |
| Proxy/Network Controller | Integrates optional proxy rotation via internal utility modules. |
| Safety Confirmation | Prevents accidental actions using rule-based match filters. |
| Result Export | Stores summaries of unmatched profiles in JSON and CSV outputs. |
- Input or Trigger — A manual CLI command or scheduled job initiates an unmatch session.
- Core Logic — The automation framework launches Bumble, scans matches, opens profiles, and triggers the unmatch dialog.
- Output or Action — Successfully unmatched profiles are logged and written to output files.
- Other Functionalities — Optional filters skip premium users, pending chats, or recently matched profiles.
- Safety Controls — Multi-step confirmation logic, UI validation, and rollback safeguards prevent incorrect actions.
Language: Python Frameworks: Appilot, UI Automator, lightweight job scheduler Tools: ADB-less control, device orchestrators, proxy utilities, internal loggers Infrastructure: Local devices, cloud device farms, containerized workers
automation-bot/
├── src/
│ ├── main.py
│ ├── automation/
│ │ ├── tasks.py
│ │ ├── scheduler.py
│ │ └── utils/
│ │ ├── logger.py
│ │ ├── proxy_manager.py
│ │ └── config_loader.py
├── config/
│ ├── settings.yaml
│ ├── credentials.env
├── logs/
│ └── activity.log
├── output/
│ ├── results.json
│ └── report.csv
├── requirements.txt
└── README.md
- Solo users use it to automatically unmatch inactive profiles so they can keep their inbox lean.
- Growth teams use it to reset devices between experiments so they can maintain consistent testing conditions.
- Automation engineers use it to validate UI flows by repeatedly performing safe unmatch operations.
- Device-farm operators use it to schedule cleanup cycles so they can ensure stable environments across dozens of phones.
Q: Does this require rooting the device? A: No, it works with standard Android automation APIs.
Q: Can it skip certain matches? A: Yes, you can configure filters for skipping based on match age, activity, or other criteria.
Q: What happens if a dialog fails to load? A: The retry logic attempts the step again with backoff and logs the failure.
Q: Can it run on multiple devices? A: Yes, it supports horizontal scaling with multiple workers.
Execution Speed: Around 18–25 unmatch actions per minute under typical device-farm network conditions. Success Rate: Approximately 93–94% across long-running sessions with automatic retries. Scalability: Supports 300–1,000 Android devices using sharded task queues and horizontally scaled workers. Resource Efficiency: Targets ~8–12% CPU and 120–180 MB RAM per worker/device depending on device speed. Error Handling: Includes structured logging, retry mechanisms, exponential backoff, alert triggers, and recovery workflows.
