Ph.D. Student in Computer Science

Mayur Akewar

Florida International University · Miami, Florida

I build systems that keep making the right decision when the evidence changes.

I am a Ph.D. student at Florida International University, advised by Prof. Janki Bhimani in the Data Management Research Laboratory. My research asks a practical systems question: how can a computational decision remain trustworthy when telemetry, workloads, and hardware behavior keep changing?

I design operational intelligence that turns noisy, fragmented evidence into grounded decisions. My work spans SSD reliability, evidence-guided agents for HPC operations, and resilient quantum-classical execution—different systems connected by the same need to reason and act under drift.

Before my Ph.D., I built machine-learning and multimodal AI systems in industry and taught computer science as an Assistant Professor. That path keeps my research close to real operators, deployable systems, and measurable failure modes.

Recent news

  1. RippleCP preprint released on counterfactual checkpoint advantage for coding agents.
  2. Completed a NERSC summer research internship and developed REPORT-OPS for evidence-grounded HPC reporting.
  3. Presented SMARTTalk at USENIX OSDI 2026 in Seattle.
  4. KORAL appeared at IEEE IPDPS 2026.
  5. WSBD appeared in the AISTATS 2026 proceedings.

Research

  1. 01GroundMake telemetry, context, and prior evidence traceable and reusable.
  2. 02PreserveRepresent temporal behavior without collapsing the trajectory into one score.
  3. 03GeneralizeTransfer operational reasoning across SSD, GPU, and QPU signals.
  4. 04ActValidate runtime decisions against current hardware state before execution.

In classical systems, I combine machine learning, large language models, knowledge graphs, and agents to turn device telemetry and operational evidence into failure predictions, explanations, and decisions. In quantum-classical systems, I study noise-aware optimization, semantic checkpointing, recovery, and resource management under calibration drift.

The unifying goal is adaptive decision-making for operational intelligence and resilient execution: systems that do not merely produce an answer, but can show what supports it, recognize when its assumptions are stale, and respond safely.

Where I apply this

AI for Systems
Machine learning and LLM reasoning for storage reliability, performance modeling, device telemetry, and operational diagnostics.
Agentic AI for HPC
Evidence-grounded agents for HPC operations, scientific computing infrastructure, and operational decision support.
Quantum Computing and Quantum ML
Noise-aware training, quantum neural networks, optimization, checkpointing, and reliable hybrid quantum-classical workflows.
Reliable and Efficient Computing
Runtime monitoring, intelligent resource management, robustness, performance, and energy-aware computing.

Publications

Automatically maintained from my Google Scholar profile; links and venue details are preserved from verified sources.

2026

2025

2024

2023

2014

A Study of Effective Load Balancing Approaches in Cloud Computing

Roshan Kotkondawar, Pushpjit Khaire, Mayur Akewar, Y. Patil

International Journal of Computer Applications, 2014

2012

A Study of Wireless Mobile Sensor Network Deployment

Mayur Akewar, Nileshsingh Thakur

International Journal of Computer Networks and Wireless Communications, 2012

Grid Based Wireless Mobile Sensor Network Deployment with Obstacle Adaptability

Mayur Akewar, Nileshsingh Thakur

International Journal of Wireless and Mobile Networks, 2012

Teaching experience, awards & academic service

1.

Teaching experience

Machine Learning · Operating Systems · Compiler Design · Cloud Computing · Web Technologies · Data Structures and Algorithms

2.

Awards

  • Best Innovative Employee Award, V2 Solutions, 2023
  • Best Instructor and Result Award, Sinhgad Institutes, 2017
3.

Academic service

  • Reviewer, Springer Nature Computer Science
  • Reviewer, Multimedia Tools and Applications

Experience

Aug 2024–Present

Graduate Research Assistant

Florida International University · Miami, Florida

Research with Prof. Janki Bhimani on intelligent systems, storage reliability, classical computing, and quantum machine learning.

Summer 2026

Research Intern

NERSC, Lawrence Berkeley National Laboratory · Berkeley, California

Developed REPORT-OPS, an evidence-guided agent for trustworthy HPC operations reporting and scientific computing infrastructure.

Apr 2021–Aug 2023

Principal Development Engineer

V2 Solutions / Spherex · Mumbai, India

Built multimodal AI pipelines and scalable systems for model training, content understanding, and deployment.

Apr 2019–Apr 2021

Senior Software Engineer

V2 Solutions / Spherex · Mumbai, India

Developed ML-based automation, media analysis, feature-extraction pipelines, and cloud services.

May 2015–Jan 2019

Assistant Professor, Computer Science

Sinhgad Institutes / Pune University · Pune, India

Taught undergraduate computing courses, mentored student projects, and contributed to departmental research.

Jan 2013–May 2015

Software Engineer

Virtual Creativities · Nagpur, India

Developed web applications and data-driven software for e-commerce and related domains.

Education

Ph.D. in Computer Science

Florida International University · 2024–Present

M.Tech. in Computer Science and Engineering

RTM Nagpur University · 2012

B.E. in Computer Technology

RTM Nagpur University · 2010

Contact

I welcome conversations about AI for infrastructure, trustworthy agentic operations, storage reliability, and resilient quantum-classical systems.