I’m Tan Ke, a second-year Data and Computer Science Master student at Heidelberg University, supervised by Sara Monji Azad. Since 2025, I have been documenting my studying and research notes on this blog. I am broadly interested in Robotics, with a current focus on world models.
I am looking for master thesis opportunities in embodied intelligence, Vision-Language-Action (VLA) models, and world models. If you are interested in collaboration or have relevant opportunities to share, please feel free to reach out by email.
On-site availability: Available to commute to the office weekly
Start date: Ready to start now for a working-student and internship role; available from August 2026 for a Master’s thesis
Expected graduation: April 2027
📑 My CV (last updated: Mar 2026) | ✉️ Letter of Recommendation (by Sara Monji Azad)
Education
M.Sc. in Data and Computer Science - Heidelberg University (Nov 2024 - Present)
Relevant coursework includes Generative Neural Networks, Computer Vision, Reasoning with LLMs, and Intelligent Systems.
B.Sc. in Computer Science - Macau University of Science and Technology (Sep 2020 - Jun 2024)
Ranked in the top 5% of the class. Final-year project: Macau Peninsula Route Planning System.
Professional Experience
AI Application Development Intern — Bihu Technology (Mar 2024 - Aug 2024)
Worked on AI-powered education products, including an LLM-driven platform for teaching-material conversion through a multi-stage prompt workflow.
Designed a LangChain-based RAG learning assistant, which was deployed at university scale and served over 5,000 students.
Data Quality Intern — Yonyou Network Technology (Apr 2023 - Jun 2023)
Supported ERP data migration and quality control by validating over one million records in SQL. Improved workflow efficiency by automating batch record generation with Python scripts.
Selected Projects
DINO-WM Representation Study
A representation study on latent world models, analyzing how different visual encoders affect action-conditioned prediction and planning under a fixed DINO-WM-style backbone.
Light Field Disparity Estimation
An ongoing light-field vision project exploring Distg-style disparity estimation. I am reproducing the original pipeline, analyzing its feature and cost-volume design, and testing an edge-guided residual refinement module to study how architectural changes improve disparity estimation quality.
Robotics Transformer (RT-1) Reproduction
A PyTorch reimplementation of RT-1 built to understand the full vision-to-action pipeline, including action tokenization, model structure, and behavioral cloning training. This project was documented in public notes as a public technical reproduction.
Diffusion Policy Reproduction
A hands-on reproduction of Diffusion Policy for visuomotor control, covering the full workflow from paper reading to implementation, training, and sampling. The project was organized as a reproducible codebase with technical notes.
RL Algorithm Implementation (PPO and SAC)
Personal implementations of Proximal Policy Optimization and Soft Actor-Critic to build a deeper understanding of policy-gradient and actor-critic reinforcement learning methods.
RagFlow
A modular FastAPI backend for Retrieval-Augmented Generation applications — a project combining practical LLM systems, retrieval pipelines, and deployable AI services.
ROS2 Basics
A structured ROS2 learning repository documenting hands-on practice with robotics software fundamentals and development workflows — I want to gain a comprehensive understanding of robotic operating systems and their development.
Macau Peninsula Intelligent Travel Route Planning System
A full-stack project for intelligent tourism route planning in Macau with Python and Streamlit. It combined attraction reviews data analysis, topic modeling, and route optimization algorithms to generate personalized travel itineraries. The project received awards in both a regional computer design competition and the undergraduate thesis evaluation.
Interests
🤖 Robotics · Artificial Intelligence
🎵 Music · Basketball · Football · Fitness · Hiking · Blog Writing
Languages
🇬🇧 English (Fluent) · 🇨🇳 Chinese (Native)