Chai

Chai is an ML-driven learning tool that integrates bite-sized tasks into your schedule, helping you achieve your goals with rich resources and notes, making learning effortless and efficient.

Comment

Inspiration

Time management is one of the most common challenges people face, particularly when trying to learn something new. Whether you’re a student attempting to master React or a working professional striving to improve, the process can often feel overwhelming. Studies indicate that 82% of people struggle with effective time management, while 70% of employees continue learning on the job. Also, 47% of high school students report that they struggle with studying. With this in mind, we discovered an opportunity to create a brilliant, ML-based solution that simplifies learning by breaking down objectives into actionable, bite-sized tasks. The solution learns to adjust to individual learning habits and professional needs, allowing users to progress gradually without getting overwhelmed. Through native integration with individual calendars, it ensures learning fits into the day-to-day, whether as a student you want to learn about courses, as an employee who is looking to change sectors, or just someone who would like to uncover new abilities.

What it does

Chai is an AI-driven productivity assistant designed to transform learning goals into structured, scheduled roadmaps. Instead of struggling to plan your learning journey manually, you can simply provide a prompt, such as "Learn React", and Chai will break it down into well-defined subtasks and milestones based on your preferred time frame and availability. Chai doesn’t just provide a list of tasks; it intelligently integrates them into your calendar, ensuring that learning fits into your daily schedule in a way that feels natural and sustainable. Each scheduled session is also enriched with curated resources, including videos, articles, and structured notes, which ML model curates from the internet, making it easier the user to stay on track. Additionally, Chai includes social progress tracking, enabling users to see how others in their network are progressing, exchange insights, and stay motivated through shared learning experiences. This tool is designed for both students and industry professionals. Whether you’re a high schooler studying American history from 1808-1914, a developer wanting to learn about LLMs, or an individual aiming to learn how to cook, Chai adapts to your needs. It also supports long-term learning goals, such as “Master basic Hindi before my trip to India,” by scheduling tasks at an optimal pace. The platform ensures that each task remains manageable, helping with consistent learning without overwhelming the user.

How we built it

Chai was developed using NextJS for the frontend, combined with Tailwind CSS and ShadCN to ensure a clean, modern, and responsive UI. For the backend, we leveraged Python Flask to handle requests and serve the application, alongside an LLM to dynamically generate structured, JSON-formatted learning roadmaps. This structured approach allows us to generate actionable steps that can be easily manipulated by our interface. We also utilized Supabase for database management, ensuring that user preferences, progress, and learning roadmaps are efficiently saved and accessed.

Challenges we ran into

Building Chai had many challenges. One of the biggest challenges we faced was with our database. At one point, we encountered unexpected failures in Supabase, which resulted in the loss of a significant portion of our stored data. This issue forced us to rethink our database structure, implement better redundancy measures, and rebuild much of what we had lost. Another major hurdle was integrating our LLM in a way that consistently produced structured and usable outputs. While the model was powerful, it did not always return results in the expected format. We had to fine-tune our prompt engineering and develop post-processing algorithms to clean and organize the output into meaningful, actionable steps for the user. This required extensive testing and iteration before we achieved a reliable system. On the frontend, we struggled to get the UI to function and appear as intended. The interface wasn’t always rendering correctly, and user interactions were not as smooth as we had envisioned. But, at the end, it all worked out.

Accomplishments that we're proud of

One of our main achievements is the robustness of Chai's ML model. We wanted to create a tool that could transform the way people approach learning and productivity, and we believe we’ve accomplished that. Chai is not just an ordinary task manager, it’s an ecosystem that actively structures learning in a way that is manageable and effective. Beyond the core functionality, we are also really proud of how adaptable Chai is. It is designed for a wide range of users, from students working through academic subjects to professionals acquiring new industry skills. This level of versatility makes the platform highly scalable, and we are excited about its potential for even broader applications. Many AI-powered tools generate vague or generic advice, but Chai is different. By leveraging structured LLM outputs, we ensure that users receive clear, well-defined steps that they can immediately act upon. Finally, we are proud of the impact Chai can have. Time management and productivity is an issue everyone faces, and we hope our platform provides a meaningful solution.

What we learned

Throughout the development of Chai, we gained valuable insights into both technical and conceptual aspects of building an AI-powered productivity assistant. On the technical side, we deepened our understanding of ReactJS and frontend development, learning how to optimize UI components for both performance and usability. We also developed better database management strategies after our experience with data loss, ensuring greater stability and reliability in our system. Working with LLMs taught us how to refine prompt engineering to achieve more structured outputs. It was a challenge to manipulate the model into providing consistently formatted data, but through trial and error, we learned how to extract the most useful information and transform it into something actionable for users. This understanding will be invaluable as we continue to improve Chai’s capabilities. Beyond the technical aspects, we also learned about the real-world applications of AI in productivity and education. AI is a powerful tool, but without the right structure, it can become overwhelming or unhelpful. Striking the balance between automation and human control was a key takeaway from this project.

What's next for Chai

Moving forward, we have many plans for Chai’s future. One of our top priorities is to enhance personalization. We want Chai to not only create learning roadmaps but also dynamically adjust them based on user progress, feedback, and personal preferences. This will allow for a more adaptive learning experience, ensuring that each user’s plan remains relevant and achievable. Integration with other productivity tools is another key area of focus. Many people rely on platforms like Notion, Slack, and Google Tasks to manage their daily responsibilities, and we want Chai to seamlessly fit into those workflows. By enabling cross-platform integration, we can make learning and productivity feel even more natural and embedded in daily routines. We are also working on developing a mobile app to bring Chai’s functionality to users on the go. Learning doesn’t always happen at a desk, and having a mobile-friendly version of Chai will allow users to access their tasks, track their progress, and engage with learning materials on the go no matter where they are. Additionally, we plan to implement direct calendar integration, allowing Chai to automatically schedule learning tasks based on a user’s availability. This will take the burden of scheduling off the user’s shoulders and ensure that learning is seamlessly incorporated into their daily life. Beyond these technical advancements, we are committed to expanding Chai’s resource library. We want to offer an even greater variety of articles, videos, and curated learning materials tailored to different learning styles and preferences. This will make learning with Chai even more effective and engaging.

Built With

Share this project:

Updates