Satyandra Guthula

Ph.D. Student, Computer Science · UC Santa Barbara

I'm a Ph.D. student in the Systems and Networking Lab at UC Santa Barbara, advised by Prof. Arpit Gupta. I work on representation learning for network data. My current focus is NetBurst, an event-centric foundation model for bursty, intermittent time series such as network traffic. Earlier, I built netFound, a foundation model pretrained on terabytes of packet traces for network security tasks. In summer 2026 I interned at Google, where I applied time-series foundation models to diagnosing faults in optical networks.

Before my Ph.D., I spent four years at VMware building the data-collection and analytics pipeline for vRealize Network Insight, a data center network monitoring product. That experience shapes how I work: I care about models that hold up on production data and survive deployment.

I'm on the job market and available from September 2027 for research scientist, ML engineer, AI engineer and ML systems engineer roles.
Photo of Satyandra Guthula

News

Research

Network data is abundant but labels are scarce, and most ML-for-networking models are trained for a single task on a single dataset. My work asks whether we can pretrain once on large volumes of unlabeled traffic and reuse the result across many tasks and networks. Current threads:

Publications & Preprints

netFound: Foundation Model for Network Security
Satyandra Guthula, Roman Beltiukov, Navya Battula, Wenbo Guo, Arpit Gupta, Inder Monga
A multi-modal hierarchical foundation model pretrained on 1 TB of raw network traces. Improves F1 by 15% over prior approaches and is resilient to shortcut learning.
Under review arXiv Code
NetBurst: Event-Centric Forecasting of Bursty, Intermittent Time Series
Satyandra Guthula, Jaber Daneshamooz, Charles Fleming, K. Wu, Walter Willinger, Arpit Gupta
A Hawkes-inspired time-series foundation model for sparse, bursty data. Improves traffic forecasting over state-of-the-art models, including Amazon's Chronos, by 2.07×–200×.
Under review arXiv
Demystifying Network Foundation Models
Roman Beltiukov, Satyandra Guthula, Wenbo Guo, Walter Willinger, Arpit Gupta
A framework for interpreting pretrained network foundation models and evaluating the quality of their representations without downstream tasks.
NeurIPS 2025 (Datasets & Benchmarks) arXiv Proceedings
Pramana: A Composable, Domain-Specific Backend for Empirical Networking Research
Jaber Daneshamooz, Eugene Vuong, Alagappan Ramanathan, Manni Moghimi, Haarika Manda, Satyam Kumar, Snithik Thode, Satyandra Guthula, Sylee Beltiukov, Dongsu Han, Tarun Mangla, Sangeetha Abdu Jyothi, Walter Willinger, Arpit Gupta
A data-generation backend that shortens the gap between forming a networking hypothesis and producing the empirical data to test it, by composing realistic bottlenecks, concurrent traffic and service-quality measurement.
HotNets 2026 arXiv
When Something Looks Too Good to Be True, It Usually Is! AI Is Causing a Credibility Crisis in Networking
Walter Willinger, Ronaldo A. Ferreira, Arpit Gupta, Roman Beltiukov, Satyandra Guthula, Lisandro Z. Granville, Arthur S. Jacobs
Argues that among the few reproducible AI/ML networking papers, even fewer produce trained models that are credible, and proposes a practical path to address this.
ACM SIGCOMM CCR 2025 Paper
Harnessing Public Code Repositories to Develop Production-Ready ML Artifacts for Networking
Punnal Ismail Khan, Satyandra Guthula, Roman Beltiukov, Roland Schmid, Tobias Bühler, Arpit Gupta, Laurent Vanbever, Walter Willinger
netMosaic, a data-collection platform that runs traffic-generating code from public repositories on programmable network infrastructure to collect data for many learning problems.
ANRW 2024 Paper
NetReplica: A Bottleneck-Regime Substrate for Reproducible Network Data Generation
Jaber Daneshamooz, Satyandra Guthula, Jessica Nguyen, William Chen, Sanjay Chandrasekaran, Ankit Gupta, Arpit Gupta, Walter Willinger
A programmable substrate for generating last-mile network data under controlled bottleneck conditions, so that ML models for networking are trained on data that reflects real bottleneck regimes.
CoNEXT 2026 Workshop (NTC-R) arXiv
Analysis of Agricultural Policy Recommendations using Multi-Agent Systems
Satyandra Guthula, Sunil Simon, Harish Karnick
Agent-based modeling of how agricultural policy affects farmers in India, from my M.Tech. work at IIT Kanpur.
arXiv 2020 arXiv

Full list on Google Scholar.

Patents (VMware)

Identification of Network Anomalies
Giri Prashanth Subramanian, Wenxuan Zhou, Satyandra Guthula, Santhosh Prabhu Muraleedhara Prabhu, Philip Brighten Godfrey
Clusters network endpoints by differences in their connectivity to identify applications and anomalous endpoints.
US 2024/0380670 A1 Google Patents
Machine Learning Based Network Anomaly Detection System
Shubhrajyoti Mohapatra, Madan Singhal, Deepak Gangwar, Satyandra Guthula
Detects anomalous network traffic from flow records serialized into temporal buckets.
US 2024/0244070 A1 Google Patents
Reinforcement-learning polling scheduler driven by correlated BGP and configuration changes across devices
Filed

Experience

UC Santa Barbara, Systems and Networking Lab
Graduate Student Researcher · Santa Barbara
2022 – present
  • Lead netFound, a multi-modal hierarchical foundation model pretrained on 1 TB of raw network traces.
  • Built NetBurst, an event-centric time-series foundation model that beats Chronos in traffic forecasting by 2.07×–200×.
  • Co-authored work on interpreting network foundation models (NeurIPS 2025) and the Pramana data-generation backend (HotNets 2026).
Google
Student Researcher · Sunnyvale (remote), Oct 2026 – present
Software Engineering Intern · Bengaluru, Summer 2026
2026 – present
  • Now extending the diagnosis pipeline with physical fault localization, using network topology for spatial context.
  • Built a retrieval-based pipeline for diagnosing optical-network faults from time series, using time-series foundation model embeddings and clustering.
  • Sent only low-confidence cases to an LLM, which made continuous monitoring practical at much lower latency and cost.
ESnet, Lawrence Berkeley National Lab
Research Intern · Berkeley
Summers 2024, 2025
  • Built a time-series foundation model for sparse, heavy-tailed IPFIX traffic: 10× lower MSE than Chronos and 94% application-classification accuracy.
  • Deployed netFound on ESnet production data; distillation halved parameters at under 5% accuracy loss. Scaled training on NERSC with PyTorch DDP and Spark.
VMware (now Broadcom)
Member of Technical Staff 2, vRealize Network Insight · Bengaluru
2018 – 2022
  • Owned the network domain of vRNI: data collection from switches, routers and firewalls, and VM-to-VM path stitching. Mentored four new graduates.
  • Redesigned the telemetry pollers, cutting memory from 12 GB to 1 GB and raising throughput 8×.
  • Built an in-memory emulator of the product pipeline that became a CI gate on every pull request.

Education

University of California, Santa Barbara
Ph.D. in Computer Science · advisor: Arpit Gupta
2022 – 2027 (expected)
Indian Institute of Technology Kanpur
M.Tech. in Computer Science · advisors: Harish Karnick and Sunil Simon
2016 – 2018
University of Mumbai
B.E. in Computer Engineering
2012 – 2016

Contact

satyandra@ucsb.edu
Department of Computer Science, Harold Frank Hall 5112
University of California, Santa Barbara, CA 93106