Building and shipping AI systems
AI / MACHINE LEARNING ENGINEER

Turning data into
useful intelligence.

I'm Aayush Oli. I’m a CSIT undergraduate at Amrit Campus focused on applied machine learning, computer vision, and backend engineering. I build end-to-end systems—from data preparation and model evaluation to API development, Docker, and AWS deployment.

SELECTED WORK

Projects with a purpose.

A selection of machine learning and software projects. Open each repository for source code, implementation details, and available documentation.

4 projects

PROJECT / 01⌁

Nepali Cultural Dress Recognition

A computer vision project for recognizing Nepali cultural ornaments and dress categories, using a ResNet50 classifier and an API-oriented deployment workflow.

PyTorchResNet50Computer VisionDocker
View repository ↗
PROJECT / 02✳

Plant Disease Detection API

A leaf-image classification service built with a ResNet18 model and FastAPI, with health checks and interactive API documentation.

PyTorchResNet18FastAPIAWS
View repository ↗
PROJECT / 03◈

AgentGuard

An AI-focused software project. Explore the repository for the current implementation, features, and technical decisions.

AI EngineeringPythonOpen Source
View repository ↗
PROJECT / 04⌘

AI / ML Portfolio & Experiments

A growing collection of machine learning practice, experiments, and implementation work across the AI/ML development lifecycle.

Machine LearningDeep LearningPython
View repository ↗
A LITTLE ABOUT ME

Curious by nature.
Practical by design.

I'm an aspiring AI/ML Engineer focused on building a strong foundation in machine learning, deep learning, computer vision, and production-minded software development.

I enjoy understanding how systems work end to end: preparing data, training and evaluating models, creating APIs, containerizing applications, and deploying services to the cloud. I value clear documentation, reproducible work, testing, and continuous learning.

↗ Building projects that solve understandable problems
↗ Learning model evaluation beyond accuracy alone
↗ Turning prototypes into documented, testable APIs
EDUCATION & ENGINEERING

Strong foundations. Real implementation.

My background combines formal computer science study with hands-on experience building and deploying complete ML applications.

EDUCATION / CURRENT

B.Sc. Computer Science & Information Technology

Amrit Campus · Tribhuvan University, Kathmandu

Currently pursuing my undergraduate degree. Relevant coursework includes Artificial Intelligence, Data Structures & Algorithms, Operating Systems, Database Management Systems, Advanced Programming, and Statistics.

APPLIED MACHINE LEARNING

From dataset to evaluated model

Hands-on work with PyTorch, CNN image classification, transfer learning, ResNet50 fine-tuning, image preprocessing, class-imbalance considerations, and evaluation using accuracy and F1 scores.

BACKEND & CLOUD

Deployable AI applications

Building REST APIs with FastAPI, documenting endpoints with OpenAPI/Swagger, packaging applications with Docker, loading models from Amazon S3, deploying services on AWS EC2, and using GitHub Actions for CI/CD workflows.

24classes in cultural-dress model
82.29%reported test accuracy · V3 model
70.90%reported macro F1 · V3 model
TOOLS & TECHNOLOGIES

My working toolkit.

Technologies I use across learning projects and application development. Specific tools vary by project.

Languages

Python · SQL · JavaScript basics

Machine Learning

PyTorch · scikit-learn · CNNs · Transfer Learning

Computer Vision

Image classification · Data preparation · Model evaluation

Backend & APIs

FastAPI · REST APIs · OpenAPI / Swagger

Deployment

Docker · Docker Compose · AWS · Render

Developer Tools

Git · GitHub · VS Code · Linux

Engineering Practice

pytest · CI workflows · Documentation

Currently Exploring

LLM applications · RAG · Agentic workflows

LET'S CONNECT

Have a project or opportunity?

I'm interested in AI/ML projects, learning opportunities, and internship conversations. Feel free to reach out or explore my work on GitHub.