I think companies in this "data analytics platform but with LLM integration" space are going to struggle in the next few years. The pain points of analytics are completely different than traditional software engineering, and things that LLMs help the most with (writing boilerplate code, bouncing between different frameworks, handling tests and docs) aren't serious issues in analytics work. I use copilot religiously for webdev work but find it completely useless for analytics.
I'm imagining something like an GPT-4 agent that has connections to your Databricks, Snowflake, Salesforce, etc.
You ask it a question and then it iteratively queries your data and produces some an answer/some code/ a graph. Or it generates a plan that you can edit a bit and then it executes.
That sounds like it would be pretty useful copilot for me? I'm not very familiar with the intricacies of analytics work though.
