Legal Tech for Healthcare: Regulatory Compliance monitoring using LLMs and Random Forests
Apply to this project here
- Sponsored by: Startup CertHub and the TUM Heinz Nixdorf Chair of Biomedical Electronics
- Project lead: Dr. Ricardo Acevedo Cabra
- Scientific lead: TUM-DI-LAB Alumni M.Sc Nicolas Gehring, M.Sc. Leon Kobinger
- TUM co-mentor: TBA
- Term: Winter semester 2024
- Application deadline: Sunday 21.07.2024
Motivation:
The Medtech industry has a regulatory and digitalisation problem: the law governing the approval of medical devices in Europe, the Medical Device Regulation (MDR), changed in 2017. All medical device manufacturers will have to adapt their processes and products to the new, much more complex regulations by 2028 at the latest. Currently, 1 in 2 companies are planning to discontinue products due to the €31 billion additional annual cost of the new regulation.
In order to survive as a European MedTech and deliver lifesaving medical products faster to patients again, MedTech companies need to digitalize and automate their documentation and processes by leveraging Machine Learning and Generative AI technologies to be able to launch new products and keep existing products on the market. Our software CertHub enables exactly this to deliver medical products years faster to patients by automating regulatory processes.
Goals:
The main objective of the DI Lab Project is to design and implement a Regulatory Monitoring functionality to keep users up to date with changing regulations. The monitoring module regularly checks publicly and publicly available data sets (e.g. EUR-Lex, PubMed, EUDAMED) on laws, standards and guidelines and abstracts them with actionable suggestions and change summaries.
Requirements for students:
Softskills: We are searching for highly motivated, innovation-minded students who prefer to work in a structured and proactive way, who have a growth mindset and an intrinsic motivation for learning.
Hardskills: Experience in the following TechStack:
DataBases: MongoDB; Analytics and Backend: Python and Django; Frontend: Next.Js and Vercel
Experience in either LangChain, Web Crawler or Random Forests
Apply to this project here