About Mandy
Mandy Alevra holds a Bachelor’s and Master’s in Management Science and Engineering from Stanford University. Her work focuses on building and deploying AI systems in complex environments. She is Chief of Staff at a frontier AI investment firm, where she works with early-stage companies on product architecture, technical workflows, and investment decisions. Her background includes work in nuclear energy and critical mineral supply chains, where she briefed federal agencies including the White House, and collaborated with technical partners across industry. Her approach is to translate complex constraints into systems that can be built and deployed.
Global Asset Capital
2025 – Current
Chief of Staff to Riaz Valani at a frontier Al venture capital firm.
Partners with early-stage teams on product architecture and deployment, and supports investment decisions through technical diligence. Experience includes building and evaluating Al systems that retrieve external data, coordinate actions within workflows, and integrate multimodal and human data.
In-Q-Tel x Stanford H4D
2025
Project Lead, Stanford Hacking for Defense Project sponsored by In-Q-Tel.
Built an AI-assisted workflow to synthesize 200+ expert interviews into a structured model of critical mineral supply chains. Modeled tradeoffs across key minerals to identify bottlenecks and optimize for resilience and domestic processing. Delivered executive briefings to federal agencies including the Department of Commerce.
Stanford Gordian Knot Center for National Security Innovation
2024
National Security Innovation Scholar at the Stanford Gordian Knot Center, working with the National Security Council.
Analyzed nuclear SMR deployment by structuring 100+ expert interviews into financial, regulatory, and export constraints. Identified bottlenecks and proposed de-risking strategies. Briefed senior NSC officials at the White House (December 2024).
Rhombus Power
2023
AI Project Intern training “Guardian.”
Labeled and transformed unstructured data into structured inputs for Guardian, an AI-driven resourcing intelligence platform for defense. Produced high-quality data for model training and evaluation.