17.12.2025

Daniil Filienko’s deep dive into privacy-preserving NLP - and what he gained from it.

Photo: Daniil Filienko

When Daniil Filienko arrived in Bochum in September 2025, he brought with him a clear research ambition and a great deal of curiosity. As a PhD student at the University of Washington Tacoma, he had long followed the work of Prof. Ivan Habernal and the Trustworthy Human Language Technologies group. The idea of collaborating with one of the leading researchers in privacy-preserving NLP felt almost out of reach at first - so much so that when the acceptance email arrived, Daniil wondered if it was real.

But from his first day at RC Trust, Daniil found exactly what he had hoped for: a focused, stimulating, and collaborative research environment where he could immerse himself fully in a single project. Supported closely by Prof. Habernal and embedded in the group’s weekly activities, he began developing a privacy-preserving synthetic text generation pipeline - a topic that connects his interests in differential privacy, membership inference attacks, and broader questions around trustworthy AI.

“I was amazed by how much freedom the German research environment offers,” he says. “I was encouraged to explore new ideas, read widely, and take ownership of the project.”

His uncertainties before traveling - especially regarding the language barrier - quickly faded. With help from team assistant Laura Knittig-Christophliemk and the RUB Welcome Center, practical matters like housing were resolved with ease. The group’s openness and international character did the rest. “I connected with researchers from different backgrounds, both inside and outside RC Trust. It truly expanded my professional network.”

But what shaped the experience most were the one-on-one meetings with his mentor. "We explored ideas deeply - sometimes very technical ones about privacy-enhancing technologies, and sometimes big foundational questions like: What do we actually mean by privacy?” These conversations, combined with informal team lunches, created an atmosphere of intellectual rigor and human connection.

Reflecting on his stay, Daniil describes the internship as an opportunity to step away from everyday responsibilities and focus intensely on one meaningful project. “Coming here was like going on an extended research retreat,” he says. “It strengthened my interest in privacy-preserving ML and deepened my understanding of the field.”

His advice to other students?

“If you’re interested in trustworthy or privacy-preserving AI, this program gives you everything you need - time, computational resources, and a supportive, collaborative environment.”

Interested in joining the RC Trust Internship Program?

Learn more and apply on the Internship website.

Category

  • Network
  • Staff
Scroll To Top