AI

Melog : AI Assistant System

This system is designed to offer holistic health management for company by integrating technology and personalized insights.

Durations:

1 year

Team size:

7

Requirements Definition

Basic Design

Detail Design

Development

Testing

Release

Maintenance

Main Features:



  • Pre-Disease Score: evaluate and monitor pre-disease conditions through scoring.

  • Condition Calendar: track and visualize health status over time with a calendar interface.

  • Rest Recommendation Days: suggest optimal rest days based on user health data.

  • Nutritional Advice: provide tailored dietary recommendations to improve overall health.

  • Billing Overview: display and manage a comprehensive list of invoices.

  • Pre-Disease Questionnaire Information: collect and analyze questionnaire data for personalized insights.

  • GPT Integration: leverage AI capabilities for advanced recommendations such as refresh days and nutrition advices to users

Devices

.

Browser(Chrome, Safari)

Server and others

.

AWS EC2, Cloud Front, Route53, Load Balancer, RDS, WAF, Cloud Watch

Technology

Laravel Laravel
VueJS VueJS
AWS AWS
ChatGPT ChatGPT
Python Python
Stripe Stripe

Challenge

Condition Calendar: Track and visualize health status over time with a calendar interface.
・The UI/UX must be highly intuitive and easy to understand.
・Users may not consistently record their data.
・Large amounts of data may slow down display performance.

GPT Integration: Leverage AI capabilities for advanced recommendations such as rest days and nutrition advice.
・RAG data is too big (3GB)
・Protecting user privacy is critical.
・Slow response times can negatively impact user experience.

Solution

Condition Calendar: Track and visualize health status over time with a calendar interface.
・Use color coding and icons to create an intuitive design.
・Implement gamification (badges, points) to encourage user engagement.
・Optimize data processing to ensure a lightweight and responsive display.

GPT Integration: Leverage AI capabilities for advanced recommendations such as rest days and nutrition advice.
・Check all documents to remove unnecessary knowledge to reduce tokens imported into the GPT Vector.
・Use anonymized data to ensure user privacy.
・Implement caching mechanisms to reduce response time.

Project Gallery

(4)

Testimonial

Mr Kim Ji Soo

COO - Bean Bros Inc

I am very satisfied with Urani's handling of the large workload. Thank you very much.