Start-Up

Buddyhere is a Berlin-based start-up developing a web platform for university buddy programs and student communities. Our mission is to help students, particularly international students, build connections and settle into university life. The platform combines registration, profiles, matching suggestions, messaging, meetings and events. University coordinators can manage participants, communicate with their communities and generate reports in one place. We are also developing Larkup as an integrated AI module within Buddyhere, enabling users to ask questions about approved information from documents and videos. The core platform features have been implemented, while the AI module is at the MVP stage and undergoing internal testing. 

Our priorities for Summer 2027 include improving the platform, developing student-facing communication and preparing events for the winter semester 2027/28. We also plan to investigate international markets and additional applications for Buddyhere. Interns will work with our team on practical projects combining product development, student engagement and business development.

This is one technical placement with two possible areas of focus: full-stack web development or AI engineering. The focus will be agreed upon according to the student’s interests, skills, and the team’s priorities. Applicants are not expected to specialize in both areas.

Full-stack focus: The intern will contribute to selected platform features, such as registration, matching, messaging, events, or coordinator dashboards. Tasks may include implementing small improvements, reproducing and fixing bugs, improving API interactions and error handling, writing automated tests, and documenting changes. The aim is to improve the usability and stability of existing workflows.

AI engineering focus: The intern will contribute to Larkup, the AI module within Buddyhere. Tasks may include preparing approved document and video test materials, improving parts of the indexing and retrieval workflow, building evaluation questions, checking whether answers are supported by sources, and measuring response times and processing costs. They may also support integration of AI responses and source references into the platform.

Work will be scoped and reviewed by the responsible team member. Expected outputs are a manageable set of tested improvements, supporting documentation, and a short demonstration or evaluation report.

Subject areas and working fields:

We welcome advanced bachelor’s and master’s students in computer science, software engineering, artificial intelligence, data science, or a related field.

For the full-stack focus, candidates should have practical familiarity with web development, including HTML, CSS, JavaScript or TypeScript, and a basic understanding of backend development, APIs, and databases.

For the AI engineering focus, candidates should have practical Python skills and a basic understanding of machine learning or information retrieval. Familiarity with language models, embeddings or retrieval-augmented generation is helpful but not essential.

For either focus, we value structured problem-solving, readable code, willingness to test and document work, and openness to feedback. Familiarity with Git is helpful. Professional experience is not required; university and personal projects are welcome. English is sufficient, and German is not required.

Review Buddyhere’s website and refresh Git fundamentals. Depending on the selected focus, review basic web application architecture and API usage, or introductory concepts in retrieval-augmented generation and AI evaluation. The technical environment and relevant codebase will be introduced during onboarding. No separate project needs to be completed beforehand.