Hari Devanathan

Hari Devanathan

Data Science

10 stories

An overview of the RAG pipeline. For documents storage: input documents -> text chunks -> encoder model -> vector database. For LLM prompting: User question -> encoder model -> vector database -> top-k relevant chunks -> generator LLM model. The LLM then answers the question with the retrieved context.
Hari Devanathan

Hari Devanathan

Health and Fitness

3 stories

Hari Devanathan

Hari Devanathan

Side Hustle

6 stories

two women in pastel pink futuristic suits and big aviator style glasses sit in a futuristic hallway and look into the distance. I like this photo to illustrate alternative ways to use ChatGPT because it evokes futurism and also not writing.
Hari Devanathan

Hari Devanathan

Investing

2 stories

Hari Devanathan

Hari Devanathan

SQL

1 story

Hari Devanathan

Hari Devanathan

AWS Cloud Engineering

6 stories

Hari Devanathan

Hari Devanathan

Personal Dev

2 stories

Hari Devanathan

Hari Devanathan

Data Engineering

1 story

Hari Devanathan

Hari Devanathan

Humor

1 story

Typewriter with a document in the feeder stating “News”
Hari Devanathan

Hari Devanathan

Real estate

No stories

Hari Devanathan

Hari Devanathan

Coding practice

1 story