Adding GSoC ideas

portal/content/articles/get-involved.md

@@ -16,7 +16,7 @@ By working with KM3NeT, you will:
 - Be mentored by experts in physics, software engineering, and data science.
 - Enhance your portfolio and contribute to groundbreaking research.
 
-We welcome students with backgrounds in physics, computer science, engineering, or related fields. Below are some example projects that you could be involved in:
+We welcome students with backgrounds in physics, computer science, engineering, or related fields. Below are some example projects that you could be involved in. For concrete project ideas for the Google Summer of Code, see tables below the areas of software, but feel invited to also make your own project proposals!
 
 ### 1. **Machine Learning for Neutrino Event Classification**  
 Develop and train machine learning models to distinguish neutrino events from background noise in KM3NeT data.
@@ -30,12 +30,29 @@ Improve the efficiency of Monte Carlo simulations used for detector performance
 ### 4. **Data Monitoring and visualization**  
 Develop web-based dashboards and monitoring tools for real-time data acquisition and analysis.
 
-### 5. **Extending KM3NeT Software Documentation**  
+### 5. **Extending KM3NeT System and Software Documentation**  
 Improve the documentation of KM3NeT’s software frameworks to make it more accessible for new contributors.
 
-### 5. **Creating LLM-based interface tools for open science**  
+| **GSoC idea**                   | KM3NeT software and systems monitoring |
+|---------------------------------|----------------------------------------------------|
+| **Description**                 | KM3NeT operates diverse systems and manages a large set of different software tools. Create an effective way to monitor software developments and system availabilities by integrating already existing tools. The technical approach to this task can be chosen freely. |
+| **Expected Outcomes**           | Monitoring dashboard, unified access to software documentation. |
+| **Required Skills**             | software management |
+| **Mentor**                      | J. Schnabel, FAU Erlangen, Germany |
+| **Project Size & Type**         | Small (90 hours) or Medium (175 hours) possible, medium difficulty. |
+
+
+### 5. **Creating LLM-based interface tools for KM3NeT**  
 Use the power of large-language-models to build interoperable scientific workflows.
 
+| **GSoC idea**                   | LLM client for KM3NeT open data products |
+|---------------------------------|----------------------------------------------------|
+| **Description**                 | Based on the [LLMTuner package](https://git.km3net.de/jschnabel/llmtuner), create a RAG-enhanced chat client for the KM3NeT open data products, and provide a python API for integration of the chat in scientific workflows. Define quality parameters for the chat client performance. Can be extended to include additional finetuning or tools application in the chat client development. |
+| **Expected Outcomes**           | RAG-enhanced chat interface, integration into KM3NeT analysis framework.  |
+| **Required Skills**             | Python, LLMs |
+| **Mentor**                      | J. Schnabel, FAU Erlangen, Germany |
+| **Project Size & Type**         | Medium (175 hours) or Large (350) possible, medium difficulty. |
+
 ## Get in Touch
 If you have any questions or want to discuss your project ideas, feel free to contact us at [opendata@km3net.de](mailto:opendata@km3net.de). We also encourage you to explore our Software or follow us on social media: