Have you ever wondered why AI tools like Windsurf and Cursor are changing the game for web development, but we’re still missing out on similar magic for data science workflows? I recently tried Windsurf and was blown away by how it can implement code for you, boosting productivity by a huge margin.
As a data scientist, I couldn’t help but think – why can’t we have something similar for our workflow? Imagine being able to prompt an agent to ‘make a predictive model to predict customer churn’ and having it execute each cell one by one, making plots, studying data, modifying outliers, and adjusting the plan as it goes before finally making a few models and testing them. That’s the standard data science workflow, but with AI assistance.
The current state of AI tools for data science is limited, with tools like ChatGPT giving generic responses that often result in errors. We need an agent that can see the outputs of our graphs, work seamlessly with notebooks, and utilize powerful GPUs on platforms like Google Collab.
I’d love to build a tool that bridges this gap, and I’m not alone. If you’re interested in collaborating or would use such a tool in your work, let me know. Who knows, maybe we can create something that revolutionizes the data science workflow.
The possibilities are endless, and it’s time we bring AI assistance to data science.
What do you think? Would you use an AI agent to simplify your data science workflow?