MIT PRIMES-USA Scholar
Massachusetts Institute of Technology | Jan 2025 – Present | Remote
During my Junior year, I participated in the MIT PRIMES-USA Program, where I worked with a graduate student, Rupert Li, on studying Friends-and-Strangers graphs. This problem is at the intersection of Graph Theory and Theoretical Computer Science. I used these graphs as a framework to prove several new theorems, including new results on random graphs. I utilized tools such as Group Theory, Graph Theory, and Probability to prove these results. You can read my paper on the arXiv: “Diameter Bounds for Friends-and-Strangers Graphs“.
My paper got accepted for presentation at the AMS’s 2026 Joint Mathematics Meetings in January ’26. Watch my presentation at the MIT PRIMES Fall Conference below.
Summer Intern
RelationalAI | Jun 2025 – Aug 2025 | Remote
I collaborated with the ML, LLM, Simulation, and PPL teams at RelationalAI to determine the effectiveness of different Generative Agent-Based Modeling environments. I also wrote code integrating tools such as Selenium, Optical Character Recognition, and Claude to extract and generate a summary of useful content from unstructured, multi-media sources such as event and conference content. I used my code to generate summaries of the 2025 ICLR and ICML conferences, spanning 2000+ research papers and presentations. These summaries were integrated with Relational’s knowledge graph to be used for future AI training.
Artificial Intelligence for Math (AI4Math) Intern
The University of Texas at Austin | Jun 2024 – Nov 2024 | Hybrid
During my internship in the Visual Informatics Lab under the guidance of Prof. Atlas Wang, I collaborated with doctorate and postdoctorate researchers to advance the use of artificial intelligence in mathematics. My work focused on developing a neural network that significantly improved the categorization of competition math problems, achieving a 20% increase in accuracy.
Building on this, I utilized the neural network to enhance the performance of a Large-Language Model (LLM) by prompting it with problem-specific strategies. This approach resulted in a 67% improvement in the LLM’s problem-solving accuracy. The outcomes of my work were documented in a single-author research paper titled “Improving Math Problem Solving in Large-Language Models Through Categorization and Strategy Tailoring.“ I presented this paper in AICCONF in June 2025.
Robotics Intern
The University of Texas at Austin | Jun 2023 – Aug 2023 | Hybrid
As a Robotics Intern in the Autonomous Mobile Robotics Laboratory, led by Prof. Joydeep Biswas, I contributed to cutting-edge research in autonomous systems. Collaborating with graduate students, my work focused on refining the accuracy of 3D campus maps generated from LiDAR point clouds. By developing algorithms in Python to process and mitigate noise in the data, I significantly reduced error rates and improved the precision of the mapping process.
This experience allowed me to engage deeply with advanced robotics concepts while applying practical problem-solving skills to real-world challenges. The internship strengthened my expertise in data processing, algorithm development, and the application of robotics in dynamic environments.
Amogh Akella 