Sarah Wassermann

Hi, welcome on my website!

I am a senior AI/ML research scientist at the intersection of Machine Learning, Natural Language Processing, and Cybersecurity. My current work focuses on leveraging Generative AI and advanced analytics to build scalable, intelligent security solutions, and production-grade AI systems.

🎯

I am currently open to challenging full remote or Zurich-based (CH) roles where I can bridge the gap between cutting-edge AI research and large-scale production systems.

🎯

As a research scientist at Proofpoint (formerly Hornetsecurity/Vade), I bridged the gap between state-of-the-art research and robust engineering. As a Product Owner, I designed an end-to-end RAG-based LLM chatbot for cybersecurity training that scales to entire companies with SLOs compatible with interactive use. I also lead a critical data-clustering project to automate threat detection, pioneered Explainable AI (XAI) initiatives to make our ML models transparent for end-users, and built the complete ML pipeline for e-mail classification.

Prior to joining Proofpoint, I was a research engineer at Huawei, where I engineered an end-to-end ML pipeline for hardware-failure prediction in hyperscale data centers using SATA/SAS SMART attributes. I was recognized with the Rising Star Award 2021 🏆.

Curriculum Vitae

Learn about what I do Get in touch with me

Research and Academic Background

My primary research interests lie in AI and computer networks (mostly in network traffic measurements and cybersecurity).

I earned my PhD. from TU Wien in 2022, focusing on ML applied to encrypted-traffic analysis and to Quality of Experience (QoE). Supervised by Dr. Pedro Casas (AIT Austria) and Prof. Dr. Tanja Zseby (TU Wien), I mainly built algorithms to monitor web and video QoE. I am still deeply driven by QoE because poor network performance shouldn't hinder our daily digital lives, which leads to all-too-common frustrations like stalling videos or slow-loading websites. I also developed an active-learning system coupled with reinforcement learning to significantly reduce the effort of manual sample labeling.

My interest in research was sparked at the University of Liège (BSc. and MSc.) and during various international research stays. Under Dr. Pedro Casas and Prof. Dr. Benoit Donnet, my MSc. research spanned Internet-path dynamics, network ML, cellular anycast, and smartphone malware. My Master’s thesis, Anycast-based DNS in Mobile Networks, was jointly supervised by Prof. Dr. Fabián Bustamante and Prof. Dr. Benoit Donnet.

To date, I have published over 30 papers in top venues like SIGCOMM, NeurIPS, and IMC (500+ citations). I also regularly serve as a TPC member and reviewer for top conferences and journals. I also maintain a strong commitment to open-source, having released RAL (reinforced active learning), NETPerfTrace, and DisNETPerf, alongside contributing to Paris Traceroute.

RAL

Reinforced stream-based active learning. RAL is an active-learning technique relying on reinforcement-learning principles, using rewards and bandit algorithms for efficiently retrieving valuable but expensive ground-truth data.

GitHub repo

NETPerfTrace

An Internet path tracking system. NETPerfTrace is a tool capable of forecasting path changes and path latency variations. For more information about this tool, please have a look at my papers.

GitHub repo

DisNETPerf

A distributed Internet paths performance analyzer developed in the context of my research internship at FTW Vienna in 2015. For more information about this work, please have a look at my papers.

GitHub repo

libparistraceroute

During my internship at LIP6 (Sorbonne Université) in Paris in 2014, I implemented Paris Ping, a generic ping tool based on libparistraceroute which can handle IPv4, IPv6 and TCP, UDP, ICMP probes. I also extended the library itself.

GitHub repo

Beyond Code

When I’m not experimenting with LLMs or analyzing email streams, you can find me traveling, capturing the world through photography, reading, or diving into video games. I also have a keen interest in web and graphic design, and love spending my free time experimenting with Adobe Photoshop. Hit me up if you want to discuss about any of those topics 😃!

Publications

Theses

PhD Thesis Machine Learning for Network Traffic Monitoring and Analysis: Application to Internet QoE Assessment and Network Security
S. Wassermann
Vienna University of Technology, 2022
MSc Thesis Anycast-based DNS in Mobile Networks
S. Wassermann
University of Liège, 2017

Journal Papers

Journal Adaptive and Reinforcement Learning Approaches for Online Network Monitoring and Analysis
S. Wassermann, T. Cuvelier, P. Mulinka, P. Casas
in IEEE Transactions on Network and Service Management (TNSM), vol. 18, no. 2, pp. 1832-1849, 2021
Journal ViCrypt to the Rescue: Real-time, Machine Learning-driven Video QoE Monitoring for Encrypted Streaming Traffic
S. Wassermann, M. Seufert, P. Casas, L. Gang, K. Li
in IEEE Transactions on Network and Service Management (TNSM), vol. 17, no. 4, pp. 2007-2023, 2020
Journal Considering User Behavior in the Quality of Experience Cycle: Towards Proactive QoE-aware Traffic Management
M. Seufert, S. Wassermann, P. Casas
in IEEE Communications Letters, vol. 23, no. 7, pp. 1145-1148, 2019
Journal Unveiling Network and Service Performance Degradation in the Wild with mPlane
P. Casas, P. Fiadino, S. Wassermann, S. Traverso, A. D'Alconzo, E. Tego, F. Matera, M. Mellia
in IEEE Communications Magazine, Network Testing Series, vol. 54, no. 3, pp. 71-79, 2016

Conference Papers

Conference Fingerprinting Webpages and Smartphone Apps from Encrypted Network Traffic with WebScanner
P. Casas, N. Wehner, S. Wassermann, M. Seufert
in 27th IEEE Global Internet (GI) Symposium, Paris, France, 2022
Conference Not all Web Pages are Born the Same. Content Tailored Learning for Web QoE Inference
P. Casas, S. Wassermann, N. Wehner, M. Seufert, T. Hossfeld
in 6th IEEE International Symposium on Measurements & Networking (M&N), Padua, Italy, 2022
Conference X-Ray Goggles for the ISP: Improving in-Network Web and App QoE Monitoring with Deep Learning
P. Casas, S. Wassermann, M. Seufert, N. Wehner, O. Dinica, T. Hossfeld
in 6th IFIP Network Traffic Measurement and Analysis Conference (TMA), Enschede, The Netherlands, 2022
Conference DeepCrypt – Deep Learning for QoE Monitoring and Fingerprinting of User Actions in Adaptive Video Streaming
P. Casas, M. Seufert, S. Wassermann, B. Gardlo, N. Wehner, R. Schatz
in 8th IEEE International Conference on Network Softwarization (NetSoft), Milan, Italy, 2022
Conference Best Paper Award Mobile Web and App QoE Monitoring for ISPs - from Encrypted Traffic to Speed Index through Machine Learning
P. Casas, S. Wassermann, N. Wehner, M. Seufert, J. Schüler, T. Hossfeld
in 13th IFIP Wireless and Mobile Networking Conference (WMNC), virtual, 2021
Conference Fast-Tracked Are you on Mobile or Desktop? On the Impact of End-User Device on Web QoE Inference from Encrypted Traffic
S. Wassermann, P. Casas, Z. Ben Houidi, A. Huet, M. Seufert, N. Wehner, J. Schüler, S. Cai, H. Shi, J. Xu, T. Hoßfeld, D. Rossi
in 16th International Conference on Network and Service Management (CNSM), virtual, 2020
Conference ADAM & RAL: Adaptive Memory Learning and Reinforcement Active Learning for Network Monitoring
S. Wassermann, T. Cuvelier, P. Mulinka, P. Casas
in 15th International Conference on Network and Service Management (CNSM), Halifax, Canada, 2019 (Fast-tracked to TNSM)
Conference Best Paper Runner Up On the Analysis of YouTube QoE in Cellular Networks through in-Smartphone Measurements
S. Wassermann, P. Casas, M. Seufert, F. Wamser
in 12th IFIP Wireless and Mobile Networking Conference (WMNC), Paris, France, 2019
Conference Beauty is in the Eye of the Smartphone Holder – A Data Driven Analysis of YouTube Mobile QoE
N. Wehner, S. Wassermann, P. Casas, M. Seufert, F. Wamser
in 14th International Conference on Network and Service Management (CNSM), Rome, Italy, 2018
Conference Anycast on the Move: A Look at Mobile Anycast Performance
S. Wassermann, J. P. Rula, F. E. Bustamante, P. Casas
in Network Traffic Measurement and Analysis Conference (TMA) 2018, Vienna, Austria, 2018
Conference Award Candidate Improving QoE Prediction in Mobile Video through Machine Learning
P. Casas, S. Wassermann
in 8th International Conference on Network of the Future (NoF), London, United Kingdom, 2017

Workshop Papers

Workshop Improving Web QoE Monitoring for Encrypted Network Traffic through Time Series Modeling
N. Wehner, M. Seufert, J. Schüler, S. Wassermann, P. Casas, T. Hoßfeld
in IFIP Performance 2020 Workshops, Workshop on AI in Networks (WAIN), virtual, 2020
Workshop I See What you See: Real Time Prediction of Video Quality from Encrypted Streaming Traffic
S. Wassermann, M. Seufert, P. Casas, L. Gang, K. Li
in 4th ACM MOBICOM Workshop on QoE-based Analysis and Management of Data Communication Networks (Internet-QoE), Los Cabos, Mexico, 2019
Workshop RAL – Improving Stream-Based Active Learning by Reinforcement Learning
S. Wassermann, T. Cuvelier, P. Casas
in European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML-PKDD), Workshop on Interactive Adaptive Learning (IAL), Würzburg, Germany, 2019
Workshop Remember the Good, Forget the Bad, do it Fast: Continuous Learning over Streaming Data
P. Mulinka, S. Wassermann, G. Marín, P. Casas
in Continual Learning Workshop at NeurIPS 2018, Montreal, Canada, 2018
Workshop Machine Learning Models for YouTube QoE and User Engagement Prediction in Smartphones
S. Wassermann, N. Wehner, P. Casas
in IFIP Performance 2018 Workshops, Workshop on AI in Networks (WAIN) 2018, Toulouse, France, 2018
Workshop BIGMOMAL - Big Data Analytics for Mobile Malware Detection
S. Wassermann, P. Casas
in ACM SIGCOMM 2018 Workshop on Traffic Measurements for Cybersecurity (WTMC), Budapest, Hungary, 2018
Workshop NETPerfTrace - Predicting Internet Path Dynamics and Performance with Machine Learning
S. Wassermann, P. Casas, T. Cuvelier, B. Donnet
in ACM SIGCOMM 2017 Workshop on Big Data Analytics and Machine Learning for Data Communication (Big-DAMA), Los Angeles (CA), USA, 2017
Workshop On the Analysis of Internet Paths with DisNETPerf, a Distributed Paths Performance Analyzer
S. Wassermann, P. Casas, B. Donnet, G. Leduc, M. Mellia
in IEEE WNM, Dubai, United Arab Emirates, 2016

Extended Abstracts and Demos

Abstract / Demo How Good is your Mobile (Web) Surfing? Speed Index Inference from Encrypted Traffic
S.Wassermann, P. Casas, M. Seufert, N. Wehner, J. Schüler, T. Hossfeld
in ACM SIGCOMM 2020 Posters, Demos, and Student Research Competition, virtual, 2020
Abstract / Demo RAL – Reinforcement Active Learning for Network Traffic Monitoring and Analysis
S.Wassermann, T. Cuvelier, P. Casas
in ACM SIGCOMM 2020 Posters, Demos, and Student Research Competition, virtual, 2020
Abstract / Demo Let me Decrypt your Beauty: Real-time Prediction of Video Resolution and Bitrate for Encrypted Video Streaming
S. Wassermann, M. Seufert, P. Casas, L. Gang, K. Li
in Demonstrations of the Network Traffic Measurement and Analysis Conference (TMA) 2019, Paris, France, 2019
Abstract / Demo Demo Award Candidate Distributed Internet Paths Performance Analysis through Machine Learning
S. Wassermann, P. Casas
in Demonstrations of the Network Traffic Measurement and Analysis Conference (TMA) 2018, Vienna, Austria, 2018
Abstract / Demo Machine Learning based Prediction of Internet Path Dynamics
S.Wassermann, P. Casas, B. Donnet
in ACM CoNEXT Student Workshop, Irvine (CA), USA, 2016
Abstract / Demo Reverse Traceroute with DisNETPerf, a Distributed Internet Paths Performance Analyzer
S. Wassermann, P. Casas
in Demonstrations of the 41th Annual IEEE Conference on Local Computer Networks (LCN-Demos 2016), Dubai, United Arab Emirates, 2016
Abstract / Demo Towards DisNETPerf: a Distributed Internet Paths Performance Analyzer
S. Wassermann, P. Casas, B. Donnet
in ACM CoNEXT Student Workshop, Heidelberg, Germany, 2015

Posters and Reports

Poster Improving Stream-Based Active Learning with Reinforcement Learning
S.Wassermann, T. Cuvelier, P. Casas
WiML Workshop co-located with NeurIPS, Vancouver, Canada, 2019
Poster Decrypting Video Quality from Encrypted Streaming Traffic
S.Wassermann, P. Casas
WiML Workshop co-located with NeurIPS, Vancouver, Canada, 2019
Poster ViCrypt: Real-time, Fine-grained Prediction of Video Quality from Encrypted Streaming Traffic
S. Wassermann, M. Seufert, P. Casas
ACM Internet Measurement Conference (IMC), Amsterdam, Netherlands, 2019
Poster BIGMOMAL – Big Data Analytics for Mobile Malware Detection
S.Wassermann, P. Casas
ACM Internet Measurement Conference (IMC), London, United Kingdom, 2017
Poster Anycast on the Move – A First Look at Mobile Anycast Performance
S. Wassermann, J. P. Rula, F. E. Bustamante
ACM Internet Measurement Conference (IMC), London, United Kingdom, 2017
Technical Report Predicting Internet Path Dynamics and Performance with Machine Learning
S. Wassermann, P. Casas, T. Cuvelier, B. Donnet
AIT-Big-DAMA Tech. Rep. A3215, 2017

Talks

Talk Active Measurements for Path Performance Diagnosis with DisNETPerf, a Distributed Internet Paths Performance Analyzer
Luxembourg Internet Days, Luxembourg, Grand Duchy of Luxembourg, November 2019
Talk Decrypting QoE in an Encrypted Internet – AI to the Rescue
RIPE 79, Rotterdam, Netherlands, October 2019

Find me on...