I AM
Prashanna

A NLP research enthusiast, fullstack software engineer and,
an overall AI practioner

About me

I'm a Computer Engineer by profession, a fulltime Software Engineer at Pasifikan (previously known as Australasia) and,
an AI enthusiast. I like to build things that assist in everyday life while also understanding the science behind it.

I'm also deeply intereted in undestanding the workings of AI (called Explainable AI) and the underlyings of Large Langugae Models. My primary research interests lies around contextual understanding in LLMs and researching efficient retrieval techniques for LLMs. I also concern myself with socio-cultural analytics and addressing bias in AI models ensuring their safe and ethical use.

July 13, 2024

Demystifying word
embeddings and word2vec

word embeddings

Word embeddings are mathematical representations of words in a high-dimensional space. Word2vec is a very popular word embedding method that uses a shallow 2-layer neural network to generate high quality word embeddings.

Jan 25, 2024

The workings of
red squiggly line

red lines

Ever wondered how the red squiggly lines in MS Word or your text editor worked? How, among hundreds of thousands of words, only the nearest correct words are suggested? This article will go in details about how this works.

Menu