Skip to Content

Hi, I'm

Dinesh Chhantyal.

Building Intelligent Software

ML Engineer • Full‑Stack Developer • GDSC President • ICPC State Champion

I build intelligent, user‑focused software combining AI/ML, scalable web technologies and real product impact. With 4+ years in software engineering, applied ML and community leadership, I deliver reliable products and help foster strong tech communities.

About Me

Context

I’m a Computer Science student (Mathematics minor) at the University of Louisiana Monroe engineering AI-as-a-service platforms, high‑performance data pipelines, and research tooling that converts raw complexity into usable insight.

What I Build

I focus on performant web & data systems for learning and science. I’ve architected scalable, secure education platforms for Clamphook and Padhao, resolving 100+ issues and cutting deployment & maintenance overhead through CI/CD automation (‑65% deploy time) and observability.

Research & Data

In collaboration with Dr. Jhim Shim, I explore algorithmic number theory (Meissel–Lehmer prime counting efficiency). At the Simons Foundation I processed 1.2 TB of BlastoSPIM embryo microscopy and built a custom 4‑channel 3D CNN trained on the institute’s ultra power‑efficient HPC supercomputer for automated nucleus state classification.

Leadership

As President of the Google Developer Student Club at ULM, I’ve grown a 250+ member developer community, driving 40% year‑over‑year event participation growth. I organize workshops, coding challenges, and cross‑department collaborations while securing grants and sponsorships and earning multi‑category marketing recognition.

Technical Interests

Current interests include: distributed inference for scientific imaging, reproducible ML evaluation, algorithmic efficiency in number theory, edge deployment patterns, and educational tooling that shortens the distance between curiosity and validated learning.

Why It Matters

I like building systems that lower friction: faster deploys mean more experimentation; structured datasets unlock better models; clear learning pathways widen access. The through‑line is leverage—turning effort into repeatable, compound impact.

If you’d like to collaborate on applied ML, education infrastructure, or data tooling, I’d love to connect.

languages
  • TypeScript / JavaScript
  • Python
  • Java
web platforms
  • Next.js
  • React
  • Nest.js
  • Node.js
  • Flutter
ai ml
  • TensorFlow
  • Scikit-learn
  • Prime Counting Methods
  • Data Modeling
  • PyTorch
  • LangChain
  • Ollama
  • OpenAI API
  • RAG Pipelines
  • Vector Stores (FAISS/Pinecone)
data infra
  • PostgreSQL & MongoDB
  • AWS
  • Google Cloud
  • RESTful APIs
  • CI/CD (GitHub & GitLab)
math cs
  • Algorithms & Data Structures
  • Number Theory
  • Linear Algebra
  • Statistics
  • Discrete Structures
core ml
  • Supervised Machine Learning
  • Unsupervised Machine Learning
cs core
  • Analysis of Algorithms
  • Data Structures
  • Object-Oriented Programming
  • Network & Data Communications
math foundations
  • Calculus
  • Linear Algebra
  • Mathematical Statistics
  • Discrete Structures
  • Building AI-driven education tools
  • Exploring distributed inference for bioimaging
  • Leading developer community workshops
  • Refining number theory research manuscript

Where I've Worked

Software Engineering Intern @ Simons Foundation
Summer 2025Website
  • Built High-Throughput 3D Embryo Imaging Pipeline: Engineered an interactive PyQt + Napari visualization and processing workflow for 1.2 TB of BlastoSPIM early-stage embryo volumetric data; parallel sliding‑window tiling accelerated exploration across 600+ timepoints and 25K+ nuclei.

  • Achieved 22× Distributed Inference Speedup: Designed SLURM-based multi-GPU workflows (Ceph/S3 streaming, memory-aware loaders) spanning mixed V100/A100 nodes, cutting model inference from 45 min to 2 min and enabling rapid experiment iteration.

Lead (President) @ GDG ULM (Google campus chapter)
Aug 2024 – May 2025Website
Webmaster @ IEEE Shreveport Section
2024 - PresentWebsite
Software Developer @ University of Louisiana at Monroe
January 2023 - PresentWebsite
Software Developer @ Clamphook
October 2022 - May 2023Website
Technical Support Staff @ Baba Computer IT Solution
January 2022 - March 2022Website
Frontend Developer @ Padhao Academy
October 2021 - December 2021Website
EPCM Content Creator and Physics Tutor @ Clamphook Academy
May 2021 - July 2022Website

Blogs & Tutorials

Other Noteworthy Projects

View Complete List of Projects/Codes
NoteMeet

An AI-powered platform for seamless note-taking and meeting collaboration. Integrate with existing platforms like Google Meet and Zoom to simplify creating, organizing, and sharing notes. Effortlessly manage agendas, track action items, and collaborate with your team. Optimize workflows with tags, categories, and quick search functions. NoteMeet enhances productivity and keeps essential details accessible.

  • NextAuth
  • JWT
  • OAuth 2.0
  • Twilio Authy
  • Resend
  • Next.js Middleware
  • Express.js
  • Prisma ORM
  • DynamoDB
  • S3
  • AWS Lambda
  • Terraform
  • Socket.IO
  • Apollo Server
  • Redis
  • Selenium
Folder
TimeClock

TimeClock is an enterprise-grade employee time management system built with Next.js 14 and TypeScript, featuring advanced authentication, real-time tracking with GPS verification, and comprehensive department management. The system leverages Server Actions for real-time updates, Prisma ORM for type-safe database operations, and includes automated reporting with email notifications via Resend.

  • Next.js 14
  • TypeScript
  • PostgreSQL
  • Prisma ORM
  • TailwindCSS
  • shadcn/ui
  • Auth.js v5
  • Server Actions
  • Server Components
  • Resend Email
Folder
CheerPal

CheerPal — Spearheaded a team of 6 members in developing a reminder application. Integrated Firebase Cloud Messaging for real-time notifications and employed machine learning models for personalized product recommendations from Amazon and Walmart, utilizing content-based filtering, K-means clustering, and decision trees.

  • Next.js
  • TypeScript
  • Payment Gateway
  • Vercel
  • Flutter
  • Firebase Cloud Messaging
  • Machine Learning
  • Content-based Filtering
  • K-means Clustering
  • Decision Trees
Folder
Collatz Conjecture

Collatz Conjecture is a simple Python program that I created to test the Collatz Conjecture. The program is designed to test the conjecture for a given range of numbers. The program is created using Python and PiP. The program is a simple implementation of the Collatz Conjecture.

  • Python
  • PiP
  • Dynamic Programming
Folder
Area Under Curve: Riemann Sum

Area Under Curve: Riemann Sum is a Python program that I created to calculate the area under a curve using the Riemann Sum method. This program solves the area under a curve using three methods: left, right, and midpoint.

  • Python
  • NumPy
  • Matplotlib
  • LaTeX
Folder
Electronic Voting System

The Election System is created using the C programming language. It is a group project that I did with my friends. The project is a simple electronic voting system that allows users to vote for their candidates. The project is part of the coursework for the course C Programming at IOE, Pulchowk Campus.

  • C Programming
  • Git
  • GitHub
  • Group Project
Folder
Covid-19 Tracker

This is a Covid-19 tracker that I created using React and Firebase Hosting. The tracker uses Disease.sh API to fetch Covid-19 data. The tracker is hosted on Firebase Hosting. The tracker uses Chart.js to display charts and Leaflet to display maps.

  • React
  • Firebase Hosting
  • Material UI
  • Disease.sh API
  • Chart.js
  • Leaflet
Folder
Rookie: Discord Bot for Class Schedule Management

Rookie is a Discord bot developed with Node.js and Discord.js to assist students in managing their class schedules. Implemented during the COVID-19 pandemic, it increased schedule adherence by 50% and improved homework submissions by 30% in our classroom Discord.

  • JavaScript
  • Node.js
  • Discord.js
Movie Flower Garden Visualization

An interactive data visualization project that represents movies as unique flowers in a garden. Each flower's characteristics (petals, size, color, shape) encode different aspects of movie data like ratings, popularity, genre, and runtime. Built with D3.js, the visualization allows users to explore films through an engaging, intuitive, and aesthetically pleasing interface.

  • D3.js
  • JavaScript
  • IMDb Dataset
  • SVG

What's Next?

Get In Touch

My inbox is always open. Whether you have a question or just want to say hello, I'll try my best to get back to you! Feel free to mail me about any relevant job updates.

Mail Me