A server manager for LiveKit, built for Multimodal Learning Analytics (MMLA) applications. SyncFlow provides scalable, robust cloud infrastructure for automated MMLA deployments, making it easier to integrate real-time multimodal data collection into AI-powered computer-based learning environments.
Status: Production Ready (Beta). Core functionality is stable and usable, but expect breaking changes as the API and architecture continue to evolve.
This project has two main parts:
- Dashboard — a Next.js application used to manage the server.
- Server — an Actix Web application used to control and manage the LiveKit services.
Don't want to self-host? Sign up at syncflow.live and start using SyncFlow by bringing your own streaming (LiveKit) and storage (S3) resources.
SyncFlow is fully containerized. The only prerequisite is Docker (with Docker Compose). For deployment of SyncFlow, we have a dedicated respository providing instructions for deploying SyncFlow on any virtual machine or cloud provider: syncflow-deploy
This project is licensed under the Apache License 2.0.
If you use SyncFlow in your research, please cite:
@inproceedings{timalsina2025syncflow,
title = {SyncFlow: A Scalable Platform for Multimodal Learning Analytics},
author = {Timalsina, Umesh and Davalos Anaya, Eduardo and Sanda, Nihar and
Zhang, Yike and Horn Fonteles, Joyce and T S, Ashwin and Biswas, Gautam},
booktitle = {Proceedings of the US Research Software Engineer Conference (US-RSE'25)},
year = {2025},
address = {Philadelphia, PA},
doi = {10.5281/zenodo.17254182},
url = {https://doi.org/10.5281/zenodo.17254182}
}This work is supported by the National Science Foundation under Grant No. DRL-2112635.