Vinod Nigade

ML Performance & Efficiency @ eBay Amsterdam

vinod_nigade.png

I am Machine Learning Performance Engineer at eBay, Amsterdam. Prior to this, I was a postdoctoral researcher in the HPDC group at VU Amsterdam. I received my PhD and MSc (cum laude) from VU Amsterdam. During my PhD, I focused on developing a DL inference serving system for latency-sensitive video analytics applications. This involved designing networked systems that leveraged various modern computing platforms, including end devices, the edge, and the cloud.

I have around five years of industrial experience in computer systems, specifically in distributed and data storage systems. I am fortunate to have received several awards and recognitions, including Best ASCI PhD Thesis Award 2023, an Outstanding Paper Award at IEEE RTSS 2022, an NVIDIA hardware grant (which includes two A30 GPUs), and a full two-year scholarship for my master’s degree.


Reviewer
IEEE Transactions on Services Computing (2025), IEEE/ACM UCC INTEL4EC (2023, 2024), NWO ICT Open (2023)

Selected Publications

  1. ACM APNet
    A Little Certainty is All We Need: Discovery and Synchronization Acceleration in Battery-Free IoT
    Gaosheng Liu*, Vinod Nigade*, Henri Bal, and Lin Wang
    In ACM Asia-Pacific Workshop on Networking (APNet) , Aug 2024
  2. IEEE RTSS
    Jellyfish: Timely Inference Serving for Dynamic Edge Networks
    Vinod Nigade, Pablo Bauszat, Henri Bal, and Lin Wang
    In IEEE Real-Time Systems Symposium (RTSS) , Dec 2022
    (Artifact Evaluated)
  3. IEEE/ACM SEC
    Clownfish: Edge and Cloud Symbiosis for Video Stream Analytics
    Vinod Nigade, Lin Wang, and Henri Bal
    In IEEE/ACM Symposium on Edge Computing (SEC) , Nov 2020
    (Acceptance Rate: 21.9%)
  4. ACM EuroSys
    Dangsan: Scalable Use-After-Free Detection
    Erik Van Der Kouwe, Vinod Nigade, and Cristiano Giuffrida
    In ACM European Conference on Computer Systems (EuroSys) , Apr 2017
  1. US Patent
    Distributed Replication in Cluster Environments
    Vinod Nigade, and Mahesh Soundalgekar
    Mar 2017
    US Patent 9,600,553