We welcome all the curious students here at AIGurukul Foundation! It is suggested that, as a preliminary study, the students take a look our recent projects, team, and the codebases online.
In this repository we shall track the progress of PS1 interns on different problem statements. Expected outcomes from the internship are multifold:
- Research papers
- Product feature
- Grokking of GenAI or MLOps fundamentals
- Improved software development skills
This repository shall host the weekly/ final presentations of the students. Through this approach, it is expected that students themselves can assess their progress and be curious to explore the unexplored.
LLMs with their emergent capablities have made it possible to think of problems in very creative ways and augment human productivity and skill at unprecedented levels. The primary and foremost application of GenAI also happens to be Education where we can aim to disseminate and distill knowledge better allowing one to learn concepts in a fundamentally better way. Motivated by this, we recently built an education personalisation tool called VidyaRANG for students, which also augments the pedagogy of teachers in schools and colleges. Hence, the broad theme for this year is AI4Education. We have certain open problems that are both research and productionisation oriented. Each of these problem statement requires a team of two or three interns to work simultaneously. These problem statements are listed as follows.
The problem statements for the first edition of PS1 - Year 2025 at the AIGurukul foundation are as follows:
- MemManage: Memory management for VidyaRANG to personalise learner's experience. See Mem0 for more context.
- KnowYourAudience: Tailoring answers based on learner's qualification and age. (Know your audience? See this for more context: Concept explained in 5 levels)
- FineTune: LLM Finetuning for personalisation based on big 5 persona (KnowsLM)
- WorthYourBuck: Evaluate and improve efficacy of VidyaRANG pipeline with open source Large / Medium/ Small Language Models (Deployable on a desktop with ease!)
- GenAIOps: Implementing CI/CD workflows and developing efficient code standards for VidyaRANG development, deployment and production environments hosted on GCP. (See VidyaRANG's architecture for reference)
Looking forward to the finalised teams by the weekend (i.e 31st May, 2025), mentor pairings, and kickoff meetings.