Software Engineer

Nilotpal Kashyap

I build AI and backend systems that are designed to scale, stay reliable, and remain efficient under real-world demand.

See the Work Get in Touch
About

Hello, I'm Nilotpal

An AI and Backend Engineer focused on building scalable systems, intelligent infrastructure, and production-ready applications.

Currently at Trumio, developing backend systems, analytics infrastructure, and secure APIs for a multi-tier SaaS platform serving 1,000+ concurrent users. Previously at Intel, building a real-time deep-learning pipeline for embedded image enhancement using custom super-resolution models.

Technologies

2+

Years Experience

India

Based In

Backend & AI

Focus

Experience

Work & Education

Jul 2024 — Present

San Jose, CA · Remote

Software Engineering Intern

Trumio

Architected backend systems and data infrastructure powering a multi-tier SaaS platform — from feature access control to real-time analytics and automated reporting.

  • •Developed a modular backend with feature-flag–based access control, enabling dynamic feature toggling across multiple subscription tiers and client segments.
  • •Built a scalable analytics admin backend to monitor and visualize activity of 1,000+ concurrent users, including detailed progress tracking and behavioral insights.
  • •Implemented a cron-based reporting service generating weekly org-level reports, saving 20+ hours of manual effort per week.
  • •Designed RBAC-secured RESTful APIs for auth and data management across PostgreSQL and MongoDB.

May 2024 — Jul 2024

India · Remote

Software Engineering Intern

Intel

Designed and built an end-to-end ML pipeline for automated pixelation detection and super-resolution enhancement, using a custom super-resolution model optimized for embedded systems.

  • •Built a two-stage pipeline to detect and enhance pixelated images using traditional ML and deep learning.
  • •Engineered a stacked ensemble classifier (XGBoost, Random Forest, HistGB + MLP) achieving 99% precision and recall on Div2K and Flickr2K datasets.
  • •Developed a custom SRGAN-based super-resolution model in PyTorch, achieving 79 FPS at 6 MB model size.
  • •Extracted HOG, FFT, DCT, and histogram features with PyTorch and scikit-learn for robust pixelation detection.
Featured

Projects

Contact

Get in Touch

Nilotpal Kashyap

LinkedIn

nilotpalk2004@gmail.com

Email

NilotpalK

GitHub

© 2026 Nilotpal Kashyap. All rights reserved.