
Teanna "Anna" Sims
BSE in Data Science
University of Michigan, College of Engineering
About Me
Hi, I'm Anna and I study data science at the University of Michigan exploring different ways computers and AI can interact with the real world. Currently, I am working on two exciting projects focused on research and development in multi-agent reinforcement learning and pose estimation for real-time video analysis.
I've had a pretty interesting journey so far when it comes to my professional pursuits. I've interned as a design engineer, quantitative trader, and even managed an XR lab. I've also been lucky enough to be part of some really cool research projects, from analyzing carbon emissions in developing countries to exploring applications of deep learning in computational finance, and exoplanet characterization.
Principles I Live By
Start Before You're Ready
Perfectionism is the enemy of progress. I've learned that waiting until I feel "ready" often means missing opportunities. The most valuable growth comes from jumping in, making mistakes, and iterating quickly.
See It Through to the End
Starting projects is easy; finishing them is hard. I commit to completing what I begin, even when motivation wanes. The discipline of following through has taught me more than any class ever could.
Create More than You Consume
In a world of endless consumption, I strive to be a contributor rather than just a consumer. Whether it's code, research, or just building something cool.
Try the Hard Thing
I deliberately seek out challenges that push my limits. The problems worth solving are rarely easy, and I've found that embracing difficulty—rather than avoiding it—leads to the most rewarding breakthroughs.
Career Goal
My ultimate goal is to become a Solutions Architect for Generative AI systems. This role would combine my technical expertise in machine learning with my passion for solving complex problems and building systems that can transform how people work and create.
Research Interests
- Real-time Video Analysis
- Cooperative Multi-agent Systems
- AutoML for Predictive Modeling
What I'm Working On Now
Creating New Agent Architectures for Concordia
Summer 2025Developing and open-sourcing innovative language model agent architectures for Google DeepMind's Concordia framework to advance cooperative AI research and lower barriers to entry for researchers.
DeJaye
OngoingDeveloping an intelligent DJ system for NeurIPS 2025 Creative AI Track that processes real-time depth video through pose estimation and GNNs to interpret crowd dynamics, automatically curating Spotify playlists responsive to audience mood.
CtrlZLearn
OngoingAggregating opportunities and learning resources for aspiring SWE, MLE, and Data Scientist.
CodePath: Intermediate & Adv Technical Intervier Prep
Summer 202512 week program teaching students to ace technical interviewing and give them a preview of real-world challenges in the industry.
Projects

architecture for cooperative multi-agent systems
Part of: Google DeepMind, Google Summer of Code

VLM capable of detecting archaeological features in the Amazon
Part of: OpenAI to Z Challenge

Event Music Curation using Real-Time Video Analysis
Part of: NeurIPS 2025 Creative AI