TaskRythm : design and implementation of an AI-based intelligent task management system
Khatri, Sauhardha; Shah, Icon; Thapa, Nishan; Bhattarai, Sanjaya (2026)
2026
All rights reserved. This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-202602152948
https://urn.fi/URN:NBN:fi:amk-202602152948
Tiivistelmä
TaskRythm is a web-based task management system designed to assist users in planning and organising work more effectively. The project originated from the observation that while many existing tools facilitate task creation, they often lack support during the early stages of planning. The objective of this thesis was to explore whether artificial intelligence could assist users in transforming general ideas into structured, actionable plans.
The system was developed as a minimum viable product using modern web technologies. The frontend was implemented with React and Next.js, while the backend was built using NestJS and Prisma ORM running on Node.js. PostgreSQL served as the database, and authentication was managed through Auth0. Artificial intelligence features were integrated using Google’s Gemini model. These technologies enabled the development of a functional and flexible system within the project timeframe.
The theoretical portion of the thesis examines software architecture, backend development, database design, natural language processing, and user experience principles. The practical portion details the project's execution, work organisation, and the implementation of core features. Key functionalities include the use of artificial intelligence to generate initial project plans from brief descriptions and to refine ambiguous tasks, alongside basic analytics to track deadlines and progress.
A usability evaluation conducted with 10 participants yielded a System Usability Scale (SUS) score of 63.5. While this score falls within the "Marginal-High" acceptability range, the feedback strongly validated the AI integration, with 100% of participants rating the AI-generated suggestions as realistic and relevant.
The system was developed as a minimum viable product using modern web technologies. The frontend was implemented with React and Next.js, while the backend was built using NestJS and Prisma ORM running on Node.js. PostgreSQL served as the database, and authentication was managed through Auth0. Artificial intelligence features were integrated using Google’s Gemini model. These technologies enabled the development of a functional and flexible system within the project timeframe.
The theoretical portion of the thesis examines software architecture, backend development, database design, natural language processing, and user experience principles. The practical portion details the project's execution, work organisation, and the implementation of core features. Key functionalities include the use of artificial intelligence to generate initial project plans from brief descriptions and to refine ambiguous tasks, alongside basic analytics to track deadlines and progress.
A usability evaluation conducted with 10 participants yielded a System Usability Scale (SUS) score of 63.5. While this score falls within the "Marginal-High" acceptability range, the feedback strongly validated the AI integration, with 100% of participants rating the AI-generated suggestions as realistic and relevant.