Harsh Bansal

About

I am a seasoned technology professional with over 20 years of experience in artificial intelligence, machine learning, generative AI, deep learning, cloud computing, and data engineering. Currently, I serve as a Specialist Analytics and Machine Learning Solutions Architect at Amazon Web Services (AWS), where I help customers build scalable and responsible AI systems using cutting-edge cloud technologies.
My industry work spans the design and deployment of production-grade, data-driven systems in enterprise environments, with a focus on applied innovation and real-world impact. I regularly collaborate with cross-functional teams to develop cloud-native architectures that power intelligent applications across sectors such as healthcare, e-commerce, and cybersecurity.
Beyond my professional role, I actively contribute to the academic and research community by reviewing manuscripts for IEEE journals, including IEEE Access and the IEEE Communications Magazine. I also participate in technical workshops, peer review of AWS blogs, and thought leadership initiatives.
I welcome the opportunity to review scholarly work that explores generative AI, applied machine learning, cloud infrastructure, data analytics, and domain-specific AI innovations. I bring a practical, impact-oriented lens to manuscript evaluation, grounded in both deep technical knowledge and extensive real-world experience.

Employment

• Amazon (United States) Specialist Analytics and AI
  2022 - Present
• Union Pacific Railroad Associate Principal Solution Architect
  2020 - 2022
• Reliance Industries (India) Deputy General Manager
  2017 - 2020
• Emerson (United States) Sr Technical Lead
  2012 - 2017

Education

• Kurukshetra University B-TECH
  1997 - 2001

Projects & Funding

Projects & funding information is unavailable.

Publications (4)

• ClinPatch-AD: Token-Level Fusion of 3D MRI Patch Transformers and Clinical Scores for Subject-Wise Alzheimer’s Disease Screening
  Harsh Vardhan Bansal, Tajinder Kumar, Purushottam Sharma +1 more
  IEEE Access 2026 DOI: 10.1109/ACCESS.2026.3690047
  +Save
• A Multimodal Deep Learning Framework Using ResNet-101 and Firefly-Based Feature Selection for Early Diagnosis of Dementia and Alzheimer’s Disease
  IEEE Access 2025 DOI: 10.1109/ACCESS.2025.3621157
  +Save
• LLMCache: Layer-Wise Caching Strategies for Accelerated Reuse in Transformer Inference
  Ised 2025 13th International Conference on Intelligent Systems and Embedded Design Proceedings 2025 DOI: 10.1109/ISED67359.2025.11405274
  +Save
• LLMCache: Layer-Wise Caching Strategies for Accelerated Reuse in Transformer Inference
  Arxiv 2025 DOI: 10.48550/arXiv.2512.16843
  +Save