Taklif.AI: LLM-Powered Platform for Personalized College Assignments
Taklif.AI is an innovative platform designed to automatically create customized college assignments that reflect students' unique interests and cultural backgrounds, utilizing Large Language Models (LLMs). Unlike traditional AI educational tools that focus only on academic achievements, Taklif.AI integrates extracurricular passions through a carefully structured prompt engineering process with both input and output safeguards. Built on a serverless architecture using AWS and Next.js, the platform leverages Llama 3.3 70B as its main LLM via LiteLLM for multi-model capabilities. This strategy seeks to enhance student engagement and reduce plagiarism by crafting assignments that cater to various interests and cognitive skills.
Key facts
- Taklif.AI leverages LLMs to generate personalized assignments.
- Platform incorporates students' extracurricular interests and cultural contexts.
- Uses structured prompt engineering with input and output guardrails.
- Serverless architecture on AWS with Next.js.
- Primary LLM is Llama 3.3 70B via LiteLLM.
- Aims to reduce plagiarism and increase engagement.
- Addresses one-size-fits-all assignment problems.
- Paper published on arXiv (2605.05842).
Entities
Institutions
- arXiv
- AWS
- LiteLLM