Steven Arellano

ML systems & GPU inference engineer: Steven co‑founded Herdora in 2025 to solve slow, expensive ML inference for production workloads. Herdora’s product focuses on profiling, kernel‑level optimizations, and optimized serving stacks to reduce latency and cost—work that leverages low‑level GPU knowledge and ML systems experience. ycombinator+2

Practical systems and infra experience from top tech internships: Steven has multiple production/infrastructure internships at Google and Two Sigma (2023–2024) and an engineering stint at Sei Labs (2022), indicating strengths in shipping infra-level code, performance work, and production systems. stevenarellano

Undergraduate researcher with lab experience: As an AxLab undergraduate researcher at the University of Chicago (Mar–Nov 2024), Steven worked on lab projects and research prototypes (ML/LLM and HCI-adjacent topics), bridging academic research and practical system-building. uchicago+1

Open-source and language/standards engagement: Steven maintains public repositories (code assistants, tooling, front-end experiments) on GitHub and is named on/associated with a C++ standards discussion paper (P3703R0), suggesting interest and involvement in low‑level language and tooling topics. github+1

Student-athlete & interdisciplinary background: Steven combined CS and Economics at University of Chicago while competing in collegiate football; this background suggests discipline, teamwork, and an ability to balance rigorous academics, research, and athletics. uchicago+1