~/AI-ML-ENGINEERAvailable for work

AASHISHBHANDARI

Building ML systems end-to-end — from model development to production. RAG pipelines, deep learning, and scalable MLOps deployments.

View Projects ↓Get in Touch →

📍 Tulsipur, Nepal

Deep Learning

CNNs · Transformers · RNN/LSTM

LLMs & RAG

LangGraph · Embeddings · ChromaDB

MLOps

Docker · MLflow · Kubernetes

Yrs Learning

Certs

Projects

// About Me

Building Intelligent
Systems That Ship.

Aspiring AI/ML engineer with hands-on experience building machine learning and deep learning projects. Skilled in model development (CNNs, Transformers), RAG pipelines, predictive modeling, data analysis, and experiment design.

Proficient with ML frameworks and tools including PyTorch, HuggingFace, LangGraph, Docker, and MLflow. Passionate about applying AI/ML techniques to solve practical problems and eager to grow through real-world projects.

View my work →

Deep Learning

CNNs, Transformers, RNN/LSTM, Siamese Networks

LLMs & RAG

RAG pipelines, LangGraph, embeddings, semantic search

MLOps

Docker, Kubernetes, MLflow, ONNX, GitHub Actions

Data & Experimentation

A/B testing, EDA, experiment design, Scikit-Learn

// Technical Skills

My Tech Stack.

Deep Learning

PyTorchPyTorch LightningHuggingFaceCNNsTransformersBERT/GPTRNN/LSTMSiamese Networks

LLMs & NLP

RAG PipelinesLangGraphEmbeddingsText ClassificationInformation RetrievalSemantic SearchChromaDB

MLOps & Deploy

DockerKubernetesMLflowOptunaONNX RuntimeGitHub ActionsAzure (AZ-900)PruningQuantization

Data & Experimentation

PandasNumPySQLEDAData VisualizationA/B TestingExperiment DesignScikit-Learn

Programming

Python (Advanced)JavaScript

Backend & APIs

DjangoFastAPIREST APIsWebSocketsRedisCelery

// Featured Work

Key Projects.

01

Visual Search System

A production-grade visual search engine allowing users to upload or paste images to find visually similar products across a comprehensive catalogue.

PythonFastAPICLIPQdrantRedisDocker

→Image-to-product search using CLIP (ViT-B/32) embeddings and Qdrant vector database
→Scalable async indexing pipeline with Redis queue and background workers
→Fully containerized microservice architecture with Nginx rate limiting

Code

02

Resume Analyzer (RAG Pipeline)

Stateful RAG pipeline using LangGraph for resume parsing, retrieval, and LLM-based evaluation. Features layout-aware PDF parsing and semantic retrieval via ChromaDB.

LangGraphFastAPITransformersDockerCeleryChromaDB

→Stateful RAG pipeline with LangGraph for structured resume evaluation
→Layout-aware PDF parsing with transformer embeddings and ChromaDB retrieval
→Async processing with FastAPI + Celery for scalable deployment

Code

03

ImageNet-based Image Classifier

High-performance classifier using ResNet/VGG with Transfer Learning on ImageNet-scale data. Applied CutMix, MixUp, and Optuna-based hyperparameter search.

PyTorchTransfer LearningOptunaGPU Optimization