NOCODESK

Inspiration

As an AIML student myself, I've observed a common challenge among my peers—grasping the theoretical concepts but struggling to practically implement them. The gap between theory and practical application arises due to the abstraction provided by Python libraries, making it challenging for students to relate the theory to real-world implementation. Inspired by this, NOCODESK aims to bridge this gap and make AI and ML more accessible to learners.

What it does

NOCODESK is an innovative platform that simplifies AI and ML for learners. It provides a user-friendly interface for data preprocessing, model selection, training, and evaluation. With NOCODESK, users can upload datasets, perform data preprocessing, choose from various ML models, train and evaluate them, all with just a few clicks. It's an interactive, educational tool that empowers users to learn and apply machine learning techniques effectively.

How we built it

The NOCODESK POC is brought to life using Streamlit, a powerful Python library for creating web applications. In the Proof of Concept (POC) stage, I coded ML models and implemented them on the IBM Z Systems LinuxONE instance using the Iris and Car Price datasets. The POC allowed me to demonstrate the platform's functionalities and showcase its potential.

Challenges we ran into

Building NOCODESK, a concept with a vast scope, presented its set of challenges. Even creating a functional POC within a 24-hour hackathon was demanding. With minimal sleep and high caffeine intake, I managed to push through the challenge. Another hurdle was working alone on such an extensive project.

Accomplishments that we're proud of

I'm immensely proud of achieving a functional POC for NOCODESK within the 24-hour hackathon duration, all by myself. This achievement has significantly boosted my confidence and reaffirmed the project's potential.

What we learned

During this project, I implemented machine learning models on an IBM Z LinuxONE instance, utilizing the Iris and Car Price datasets as my testing data for the NOCODESK POC. This experience further solidified my understanding of machine learning models and data processing.

What's next for NOCODESK

The future of NOCODESK involves completing the POC on Streamlit, pitching the project to potential collaborators, and securing funding. Once funded, I'll expand the team to work on developing the full-stack website version of the platform. This website will be deployed in collaboration with my college's School of Computer Science and Engineering, serving as a testing ground for the platform. Based on feedback and results from the student community, we'll continually enhance NOCODESK and look to collaborate with more universities.