How I use AI

Preface

I want to preface this write-up by acknowledging my strengths. I have built Cars from scratch in my Undergrad and I have designed heavy-duty truck architecture from scratch as an automotive engineer. Therefore I, with reasonable confidence, can claim that am fairly well-versed with system-design in general, long before I stepped into the world of Data.

Definition

AI: Any tool that becomes a bridge between my technical shortcomings/inexperience and my ideas/execution.

Use

Idea to execution From an idea, my initial focus is to generate a technical specification and requirements document. I do this using either ChatGPT or Gemini. I then look for the datasets and data sources, explore them using python and establish an initial low level POC. Desigining high-level system design and architecture follows. I dont claim to be a seasoned System architect or designer but I understand data pipelines and workflows to create an effective product. The POC and high-level design serve as fodder for the prompts to coding agents like Claude/Gemini to help me develop and code the pipeline into existence. Strict policies and execution restrictions are a part of my process and my domain knowledge helps me make products that are better than "prompted projects".

I am aware of the discourse around this technology, both positive and negative. I acknowledge its impact on society, environment and people. However, I believe as a tool in the right hands there is potential in helping us improve and am willing to give it room to mature before I embrace it completely or reject it.