KUnnect

KUnnect (alpha) Contact

Contact Us

We'd love to hear your feedback, questions, or suggestions!

Name (optional)

Email (optional)

Type

Message *

Thank you! Your message has been received.

Data Collection Policy

We collect non-personally identifiable information (number of visits, etc.) and we use cookies to support functionality such as voting. We do not collect any personally identifiable information such as IP addresses, with the exception of data voluntarily submitted via the Contact form.

Reason for Downvote

Please provide a reason for downvoting this recommendation.

Find Collaborators

Search for a researcher to see their current and potential collaborators

Search collaborators at KU only

Upload Papers for Recommendations

Upload 1-5 research papers (PDF) to find potential KU collaborators

📄

Click to select or drag and drop PDFs here

Max 5 files, 30 MB each

This feature requires a GPU and is currently unavailable.

Loading results...

Error

Disclaimer: This app is a prototype sponsored by SUND's Vice Dean of Research, not an official KU-IT app

KUnnect v0.3 • How does it work? • Data collection policy

How does it work?

Semantic Embeddings: We use the Nomic AI embedding model to transform publication abstracts into high-dimensional vectors. This allows our system to understand the semantic meaning of research content and identify conceptual relationships between different research areas.

Retrieval-Augmented Generation (RAG): Our RAG system uses a local large language model - currently we are using gpt-oss-20B, from OpenAI - to provide intelligent ranking and explanation of search results, enabling a more intuitive and powerful search experience.

Development: Parts of the code for this application and visual elements such as the infographic above were produced with tools such as Claude Code and Gemini Pro 3. They can help dramatically speed up the time spent developing small applications such as these.