Portrait of Anni Zimina standing in front of a colorful mosaic wall at Stanford

Anni Zimina

MS Computer Science
Stanford University

Anni Zimina

Anni is a Computer Science student at Stanford Engineering focused on agentic evaluation and multimodal classification. She has built Target-Evaluate-Improve (TEI) Loop, an open-source agent evaluation framework, and a multimodal classification system classifying millions of social media profiles across thousands of parameters.

Papers

  1. 01

    AI-Driven Real-Time Persona Fusion Agents for Adaptive Multimodal Marketing Intelligence

    Presented at the 2025 Stanford Causal Science Center Conference on Experimentation

  2. 02

    TEI Loop: Structural Fixes and Pareto-Front Optimization for Agentic Systems

    Presented at AI + Science: Accelerating Discovery, Stanford HAI
  3. 03

    Confluence: A Unified-Percept Architecture for Multimodal Creator Classification at Scale