Inspiration
After the successful opening ceremony, I reviewed the hacker guide for the available projects. ServiceNow stood out as the most unique and challenging project, with well-defined tasks, clear expected outcomes, and a provided dataset to work with. The attention to detail and diligence they showed inspired me to take on the challenge and push myself
What it does
It is a recommendation system that suggests products and accelerators to both existing and new customers.
How we built it
First, I conducted extensive research on various AI and machine learning algorithms used in building recommendation systems. Next, I planned the project and divided it into different phases, such as building the backend service, designing the UI, developing the algorithms, connecting the UI to the backend, and integrating the algorithms with the backend. Finally, I performed testing using the available dataset.
Challenges we ran into
I encountered conflicts with some Python packages, where certain versions were either outdated or lacked the functions I needed. This required me to either downgrade or upgrade the packages depending on the specific use case. Additionally, I faced CORS issues when connecting the backend with the frontend. Another challenge was the limited manpower, as I was the only one working on the project.
Accomplishments that we're proud of
Stepping out of my comfort zone to take on this challenging project helped me learn about data manipulation, data sanitization, and the capabilities of data science and AI. Presenting the project also boosted my confidence in my skills.
What we learned
This project provided me with hands-on experience in key data science concepts and practical skills in building scalable recommendation systems by leveraging a combination of machine learning techniques
What's next for ServiceNow: Accelerator Recommendation Engine
With additional training and more data, the Accelerator Engine will deliver increasingly accurate and impactful recommendations to its users. Additionally, integrating a Large Language Model (LLM) will enhance user interaction, making the AI bot more interactive and engaging.
Log in or sign up for Devpost to join the conversation.