Phd/Postdoc in AI and wireless communications

Priyanka Kaswan

******@*********.*** j +1-301-***-**** j LinkedIn j Personal Website

Objective

Seeking to apply strong background in ML, backend systems, and sensor data processing to impactful software engineering, data scientist and ML engineering roles, with a PhD in ECE from University of Maryland and postdoc at Princeton University.

Technical Skills

Languages: Python, C++, Java, MATLAB, SQL

Frameworks/Tools: PyTorch, TensorFlow, Keras, Node.js, Git, Jupyter

Core Expertise: Backend development, ML model training and evaluation, LL-based inference, distributed systems, federated learning, optimization, reinforcement learning, AoI, informative bandits

Systems: Linux, AWS (basic), Docker (basic), GitHub Actions

Courses: Foundations of Deep Learning, Generative AI with Large Language Models by AWS Education and Current Appointment

Postdoctoral Research Associate, Princeton University 2024 { Present Advisor: Prof. Andrea Goldsmith

Research on impact of age of information on real-time network control and inference. Ph.D., Electrical & Computer Engineering, University of Maryland, College Park 2019 { 2024 Advisor: Prof. Sennur Ulukus GPA: 4.0

Researh on timely edge intelligence.

Awards: Charles A. Caramello Dissertation Award, ECE Distinguished Dissertation Fellowship. B.Tech, Electrical Engineering, Indian Institute of Technology (IIT), Delhi 2013 { 2017 Experience

Qualcomm Technologies Inc. { Systems Engineering Intern Summer 2023

(recommended for full-time hire)

Developed statistical AI/ML model monitoring system for 5G/6G interference power prediction.

Published a paper (IEEE ICC 2024) & led a U.S. patent on ML model update via importance weights. UrbanClap (now Urban Company) { Backend Intern Summer 2016

Built backend APIs using Node.js to enhance Search Engine Optimization for city service pages. Research Assistant, UMD / Princeton 2019 - Present

Built decentralized MoA inference framework for LLMs deployed on edge nodes using gossip protocols.

Developed rAge-k federated learning algorithm with AoI-driven selective gradient updates.

Implemented scalable data dissemination protocols using Python; achieved sublinear AoI scaling. Projects

Prune-m: Designed optimization framework for federated gradient compression.

Beamforming with MUSIC: Built speech enhancement for mobile audio.

Informative Arm Bandits: Developed methods for identifying informative arms in bandit settings. Publications & Patents

\Statistical AI/ML Model Monitoring for 5G/6G:Interference Prediction CaseStudy",IEEE ICC 2024.

US Patent Application 18/818,273: Techniques For Modifying ML Models Using Importance Weights.

\Distributed Mixture-of-Agents for Edge Inference with LLMs", submitted to ICC 2024.

\rAge-k: Federated Learning Using Age Factor", IEEE Asilomar 2024.

Multiple IEEE journal and conference papers on ML, control, wireless. Honors & Awards

Rising Stars in EECS, MIT, 2024.

Finalist, Indian National Mathematical Olympiad, 2013.

Contact this candidate