Ask HN: Seeking Feedback on an Idea to Simplify Machine Learning Model Building

[prism_wisdom]

Whether you’re a data scientist, non-tech researcher, software engineer, or someone curious about machine learning, have you ever felt overwhelmed by coding complexities, managing experiments across models, and dealing with infrastructure? I’ve often found myself spending more time managing technical hurdles than focusing on building and optimizing models.

That’s why I’m developing a platform that enables users to effortlessly build, train, and compare multiple models in parallel—without needing to code or worry about infrastructure management. The goal is to let you focus on solving problems and innovating, while the platform handles the technical complexity for you.

I’d love to get your feedback on this concept: 1. Do you think this would help simplify your workflow or boost productivity? 2. What specific features or pain points should the platform address? 3. Would you use a tool like this, or do you have any concerns about it?

We’re in the early stages of development, so your input will be invaluable in shaping the platform. I’m excited to hear your thoughts! Feel free to share your experiences or frustrations with machine learning platforms—your feedback will be instrumental in helping us create something truly impactful.

[razodactyl]

Hey. I'm keen. I solo-developed my own compact LLM and living in JupyterLab gets overwhelming.

[prism_wisdom → razodactyl]

Thanks so much for your reply! It's great to hear that you're interested. It sounds like you're already doing some heavy lifting in JupyterLab.

I’d love to hear more about what specific features or functionalities you think could make your workflow easier. For example, would support for more complex algorithms or advanced model comparison be useful? Right now, the WIP platform supports basic classification and regression algorithms, and you can run up to 3 models in parallel—but we're always looking for ways to expand its capabilities.

Any insights or ideas you have would be really helpful as we continue developing the platform! Thank you.