index / 6

project writeups

Work, notes, and things learned.

#1

Python / PyTorch / CUDA

vLLMini

A minimal inference stack for understanding how modern LLM serving systems handle model loading, quantization, attention, and token sampling.

read writeup
#2

FastAPI / Next.js / ML

Conflux

A civic intelligence workflow for turning noisy multilingual public reports into clustered, location-aware infrastructure proposals.

read writeup
#3

Python / PyTorch / NLP

VectorAlign

A low-resource translation support tool focused on extracting reliable bilingual alignments from multilingual embedding spaces.

read writeup
#4

LangChain / FastAPI / RAG

FloatChat

A natural-language interface for exploring structured oceanographic data without manually writing SQL or parsing NetCDF metadata.

read writeup
#5

PyTorch / OpenNMT / Modal

LINGO

A focused machine translation project for Kangri-Hindi that moves from dataset preparation through training and installable deployment.

read writeup
#6

Python / OpenCV / PyTorch

DeltaVision

A visual inspection prototype combining computer vision, temporal modeling, and a fast feedback UI for quality analysis.

read writeup