Linda Li

I am an enthusiastic Computer Science student passionate about AI, Data Science, and Software Engineering.

Hello! I’m Linda Li, a master student in Computer Science at TU Berlin, passionate about AI, Data Science, and Software Engineering.

🎓 Ecducation

  • M.Sc. Computer Science - Technische Universität Berlin
    Apr 2025 - Present

    • Currently in the second semester, focusing on ML/DL, LLMs, and data infrastructure.
    • Developing an AI-powered language learning application.
    • Working on an Agentic RAG project to improve enterprise knowledge retrieval.

  • B.Sc. Naturwissenschaften in der Informationsgesellschaft (Natural Sciences in the Information Society) - Technische Universität Berlin
    Oct 2020 - Oct 2025

    • 8 semesters, 240 ECTS
    • Interdisciplinary program combining Mathematics, Physics, Computer Science, and Chemistry
    • Focus on Computer Science
    • Bachelor’s Thesis: “Development of a Methods Catalog for the Application of Data Analytics in Manufacturing”
      • Accepted for publication at GQW-Tagung 2025 (2nd author, Best Paper Award)

💼 Work Experience

  • AI Research Engineering Intern – Max Planck Institute for Human Development (Center for Humans & Machines), Berlin
    Jul 2025 – Sep 2025
    • Independently developed a full-stack experimental platform (Learning to Teach project) to study AI-supported human reasoning, combining a React + MUI frontend with a FastAPI + PostgreSQL backend.
    • Designed and implemented data pipelines for capturing fine-grained user interaction logs (timestamps, AI feedback, socket-based data streaming), ensuring high data quality for behavioral experiments.
    • Built and deployed containerized services with Docker and GitLab CI/CD to Kubernetes clusters.
    • Integrated OpenAI LLMs and Raven-based fine-tuned models to provide different reasoning supports, including system prompt design and feedback optimization.
    • Developed embedding workflows for semantic analysis using Hugging Face models on Raven HPC, enabling large-scale retrieval, clustering, and evaluation of reasoning data, with GPU acceleration and Slurm job management.
    • Created a proof-of-concept chatbot (Collective Voices project) to compare fine-tuned models on Raven with OpenAI models, using Gradio for rapid prototyping and conversational evaluation.
    • Contributed to experiment design and evaluation, implementing admin dashboards and visualization tools for tracking participant progress, survey results, and system integrity checks.
    • Collaborated closely with supervisors and researchers in weekly check-ins, presenting technical progress and discussing experimental methodology.

    Technologies & Tools: React (MUI), FastAPI, PostgreSQL, Docker, Kubernetes, GitLab CI/CD, HPC (Raven, Slurm), Embeddings for Semantic Analysis (Hugging Face), Fine-tuned LLM Integration, Gradio, Socket.IO, Data Visualization Working Languages: English, German

  • Research Assistant & Tutor – Technische Universität Berlin
    Jun 2024 – Apr 2025
    • Independently taught tutorial sessions (Übung) for the course Introduction to Engineering Data Analytics with R, including conducting weekly office hours and preparing exercise materials.
    • Set up and managed the course website and student learning platform.
    • Independently developed a full-stack web application for the ReDaQ project, an intelligent data quality management platform, using Flask (Python) and integrating Rasa and LLaMA3 models for chatbot functionality.
    • Contributed to the preparation of the first two chapters of the book “Datenbasiertes Qualitätsmanagement: Mit Datenanalyse und KI-Methoden zur Digital Quality”, supporting introductory texts and illustrations.
    • Supported the MIQFEM project by conducting Docker-based web application testing and documenting system logs and technical processes.
    • Received highly positive feedback for technical expertise, problem-solving ability, reliability, and teamwork.

    Technologies & Tools: Python (Flask, Pandas, Numpy…), R, Docker, HTML/CSS, JavaScript, LLM Integration (LLaMA3), NLU model integration (Rasa), Markdown, Typo3
    Working Languages: German, English

  • Data Analytics Intern – FZI Forschungszentrum Informatik, Berlin
    May 2024 - Sept 2024
    • Conducted an independent project analyzing bus delays in Hamburg based on passenger volumes, route lengths, time periods, and weather conditions.
    • Collected, processed, and integrated real-time data from APIs and open sources.
    • Applied statistical analysis (ANOVA, Lasso regression) and time series forecasting (Prophet model) to identify delay patterns and propose optimization strategies.
    • Strengthened skills in Python programming, data engineering, and applied statistics.
    • Received highly positive feedback for technical expertise, problem-solving ability, reliability, and teamwork.

    Technologies & Tools: Python (Pandas, NumPy, Matplotlib, Scikit-learn, Prophet, Geopy…), API integration, data preprocessing, statistical modeling, Latex
    Working Languages: German

🛠️ Technical Skills

Programming Languages:

Python
R
C
Julia
SQL
Haskell
React
Java
JavaScript
HTML/CSS
Prolog
MIPS Assembly
Flutter

Tools:

LaTeX
Markdown
Figma
Git
Docker
Kubernetes

🌍 Language Skills

  • Chinese: Native
  • German: C1
  • English: B2
  • Japanese: A1
  • Spanish: A1

🚀 Projects (🔗 See full project portfolio)

Industry Projects

  • Agentic RAG – Intelligent Knowledge Retrieval and Reasoning System
    Ongoing research project under Fraunhofer FOKUS and Technische Universität Berlin, focusing on agentic retrieval-augmented generation (RAG) for reasoning over mixed-type knowledge bases that combine structured and unstructured data. The project aims to develop autonomous LLM agents capable of query understanding, knowledge grounding, and SQL reasoning to enhance transparency and interpretability in scientific and biomedical question answering.

    Currently in active development.

  • Lumin – AI-Powered Vocabulary Learning App
    Co-founded an educational startup project with two teammates, originating from the TU Berlin course Projektorientierte App- und Web-Entwicklung.
    The app integrates AI-generated exercises with personalized review mechanisms to enhance language learning efficiency. Currently in the process of seeking investment opportunities and applying for startup fellowships.

    Further details coming soon.

  • ReDaQ – Intelligent Data Quality Management Platform
    Independently developed a full-stack web application for the ReDaQ project using Flask (Python), integrating Rasa for dialogue management and LLaMA3 via API.
    Designed custom dialogue flows, crafted prompts to align with user intents, and implemented a Retrieval-Augmented Generation (RAG) pipeline to support context-aware responses.

    Demo available on GitHub.

  • Public Transport Delay Analysis (FZI Internship Project)
    Conducted a data analysis project at FZI Berlin, analyzing the impact of factors such as passenger volume, route length, time periods, and weather conditions on bus delays in Hamburg.

    Further details coming soon.

University Projects

  • 3D Sphere Texture Projection and Image Processing Project
    Developed Python modules and contributed to image processing with Julia, focusing on projecting PNG textures onto 3D spheres.

  • Virtual File System Project
    Independently developed a virtual file system in C with directory management, file read/write access, and data import/export functionalities.

  • “Death Stacks” Bot Development
    Built an intelligent game bot in Haskell, emphasizing functional programming and software quality assurance.

📍 More university project details: GitLab.

  • Academic Poster: Gekoppelte Schwingungen (Coupled Oscillations)
    Selected to create a scientific poster on coupled oscillations as an extension to the Physikalisches Grundpraktikum (Physics Laboratory Course) at TU Berlin.

    📍 Poster details