The CASP Systems lab has multiple open positions for PhD students and postdocs. We invite applications from strong candidates who are interested in scalable data stream processing, systems for large-scale graph analysis and machine learning, and systems for privacy-preserving data analytics. To apply for a PhD position, submit your materials by... More
CASP lab receives two new research awards: A gift from Bosch GmbH to support our research on privacy-preserving data processing systems. A SaTC Core Medium award from NSF (#2209194). The award will support our research on secure outsourced analytics in untrusted clouds for the next 4 years. We are grateful for... More
Our paper, "Secrecy: Secure collaborative analytics in untrusted clouds", by John Liagouris, Vasiliki Kalavri, Muhammad Faisal, and Mayank Varia, has been accepted for presentation at NSDI'23! Secrecy is a system for privacy-preserving collaborative analytics as a service. Secrecy allows multiple data holders to contribute their data towards a joint analysis in... More
The following papers were accepted for presentation at upcoming SIGMOD'22 workshops: Evaluating Model Serving Strategies over Streaming Data by Sonia Horchidan, Emmanouil Kritharakis, Vasiliki Kalavri and Paris Carbone was accepted at the 6th Workshop on Data Management for End-to-End Machine Learning (DEEM'22). GCNSplit: Bounding the State of Streaming Graph Partitioning by Michał... More
The paper “The Non-Expert Tax: Quantifying the cost of auto-scaling in Cloud-based data stream analytics.” by Yuanli Wang, Baiqing Lyu, and Vasiliki Kalavri, has been accepted for presentation at the BiDEDE’22 workshop.
Shengyao (Jax) Luo is joining the CASP Systems lab this summer, after being awarded a UROP grant. Jax's project, "Monitoring and improving energy consumption in stream data processing", aims to curate and publish a unique data set of fine-grained time-series energy and performance data for streaming workloads.
Our paper "A New Benchmark Harness for Systematic and Robust Evaluation of Streaming State Stores" has been accepted for presentation at the EuroSys'22 conference.
Our research lab has received three Research Incubation Awards from the BU Red Hat Collaboratory: 1. Towards high performance and energy efficiency in open-source stream processing (PIs: Vasiliki Kalavri, Jonathan Appavoo) 2. Serverless Streaming Graph Analytics (PI: Vasiliki Kalavri) 3. Secure cross-site analytics on OpenShift logs (PI: John Liagouris) Read the full announcement here.
Our paper "Learning on streaming graphs with experience replay" has been accepted to appear at the 2022 ACM/SIGAPP Symposium on Applied Computing (SAC'22). This is a collaboration with Massimo Perini (University of Edinburgh), Giorgia Ramponi (ETH Zurich), and Paris Carbone (KTH Royal Institute of Technology). See the preprint pdf here.
We invite applications for a postdoc position in self-managed and power-efficient stream processing systems. Applications will be reviewed on a rolling basis starting November 22 and until the position is filled. More information and instructions on how to apply can be found here.