• 

  operations

• 

  column information

• 

  final report

• 

  problem statement extraction

• 

  data set and problem statement upload

• 

  data preview

• 

  login page

• 

  first page

Inspiration

We were inspired by the growing need for automated and intelligent data solutions in a world overflowing with data. Data analysts, researchers, and business users often spend hours cleaning data, running EDA, and building models manually. We wanted to build a platform that empowers users to focus on decision-making, while automation handles the heavy lifting.

What it does

Slack AI is an AI-driven platform where users can upload datasets along with problem statements. The system automatically processes the data—cleaning it, analyzing it, and generating visual insights. It also builds appropriate machine learning models and produces a complete final report that addresses the user’s original questions.

How we built it

I used bolt.new for the full development of this project and for the API I used openrouter deepseek r1 model for the API

Challenges we ran into

Handling messy, unstructured datasets with missing or inconsistent formats.

Automating model selection without user intervention.

Designing a clean UI that simplifies a complex process.

Accomplishments that we're proud of End-to-end automation: from data upload to final report.

Seamless user experience with minimal manual input.

Integration of EDA, visualization, and ML in one place.

What we learned

We gained deep insights into building scalable AI pipelines and learned how to handle real-world dirty data. We also improved our skills in API design, UI/UX integration, and report automation.

What's next for Slack AI

We plan to support real-time data streaming, allow custom model tuning, and introduce a collaboration feature where teams can annotate and discuss insights directly on the platform. We're also exploring integrations with tools like Slack and Microsoft Teams.

Built With

• bolt.new
• openrouter