Thanks to ChatGPT, we all are adapting AI in our day-to-day life. But we are still way far in customizing this AI as per our needs to do some specific tasks at a personal or public scale.

With Sttabot, I aim to reduce this barrier by making AI accessible and tailorable to all. So, my agenda with this paper would be to let anyone of any age and any profession, to tailor & customize their AI experience and build products that were just an imagination before.

Think of this technology as a completely no-code solution that can turn your prompts (or definitions) into an AI application in few seconds.

An introduction to the problem.

Most of us are familiar with the potential of tools like ChatGPT. Even, many of us are utilizing it for automating our little day-to-day tasks like doing assignments, drafting emails, college project, gramar, doing research, etc.

However, ChatGPT is only a frontend to interact with a LLM (Large Language Model) engine that is the TurboGPT-3.5 or the GPT-4 engine.

And if have made an observation, the more you define the problem (give context to the model), the more reliable is the answer.

Let me show this to you with an example :

Here, we asked the AI to give us a list of 3 investors for our startup.

This is a classic conversational case of how most people talk to interfaces like ChatGPT.

Now, compare this with a well-written prompt that demands the same output but gives context on how the output should be presented and what should it contain to get the desired result :

Here, though we are asking the AI to give us the same results of 3 potential investor, the criteria that we defined gave a completely different result - one which is more accurate and considerable.