Terminal

$ whoami

Mauricio Tec · Sr. Applied Scientist · Amazon AGI

$ cat about.txt

Agentic RL post-training for frontier models at Amazon AGI. Previously built autonomous robots, trained RL policies for public health, and trained foundation models at Harvard.

$ open robocup.jpg

$ cat experience.txt

Amazon AGI · Harvard · Meta AI (FAIR) · Intel · CI Strategies
PhD UT Austin · MSc Cambridge University · BSc ITAM

$ ls projects/examples/

• rule-bottleneck-rl — RL agents that explain their own decisions

• robust-dpo — making RLHF work when humans disagree

• diffusion-transfer — adapting diffusion LLMs without retraining

• adversarial-intrinsic-motivation — teaching RL agents to explore

• robot-soccer-vision — real-time deep learning vision on low-end robot hardware

• heat-alerts — RL for public health policy

• equivariant-topo-nn — geometry-aware neural networks for complex structured data

$ cat links.txt

github · scholar · linkedin · x

$ cat /etc/motd

// …built with Claude Code. based on data, but verify.

$ _