Skip to content
ZZ / CIS
01

Computer & Information Science

Zijie Zhao

PhD student · Penn Database Group
Philadelphia, Pennsylvania

Portrait of Zijie Zhao
Leveraging AI14.9B lifetime tokens
Codex · 2026-09-17Less More

I build adaptive software systems that learn from execution and improve how they behave, with a current focus on database engines.

02

Current focus

Self-evolvable software

I am a PhD student in Computer and Information Science at the University of Pennsylvania, advised by Ryan Marcus. My work sits at the intersection of database systems, machine learning, and machine programming. I am interested in software that can observe its own behavior, adapt at runtime, and evolve without losing the guarantees that make systems dependable.

Advised by Ryan Marcus (opens in a new window) at the Penn Database Group (opens in a new window).

03

Active lines of work

Research threads

  1. R.01

    Adaptive execution

    How can a database choose better physical behavior at microsecond timescales?

    I study execution engines that learn from fine-grained runtime feedback and change strategy as data and workloads shift.

    Piece of CAKE ↓
  2. R.02

    Workload understanding

    What do production traces hide about the work users actually wanted to run?

    I examine the feedback loop between platforms and users so workload data can be interpreted without mistaking survivors for demand.

    Survivorship Bias ↓
  3. R.03

    Self-evolvable software

    Can software adapt inputs, policies, and implementations instead of waiting for retraining or manual tuning?

    My broader goal is to combine machine learning with program structure to build systems that improve while they run.

    CodeImprove ↓
04

Selected · reverse chronological

Publications

P—01

Piece of CAKE: Adaptive Execution Engines via Microsecond-Scale Learning

Zijie Zhao, Ryan Marcus

A microsecond-scale learning system selects a physical kernel for each data morsel, reducing end-to-end workload latency by up to 2×.

  • DATABASE SYSTEM
  • QUERY PROCESSING
  • MACHINE LEARNING
Venue: UNDER REVIEW
Access: PDF / record
P—02

Survivorship Bias in Industrial Database Workloads

Ryan Marcus, Jeffrey Tao, Peizhi Wu, Zijie Zhao · All authors contributed equally.

Shows how production traces encode a negotiation between users and platforms—and why the queries missing from those traces matter.

  • DATABASE SYSTEM
  • QUERY OPTIMIZATION
  • SURVIVORSHIP BIAS
Venue: CIDR 2026Best Paper
Access: PDF / record
P—03

CodeImprove: Program Adaptation for Deep Code Models

Ravishka Rathnasuriya, Zijie Zhao, Wei Yang

Adapts out-of-scope programs with semantics-preserving transformations, improving deep code models without frequent retraining.

  • SOFTWARE ENGINEERING
  • DEEP LEARNING
  • PROGRAM ANALYSIS
Venue: ICSE 2025
Access: PDF / record
P—04

NuClass: An ontology-driven vision–language foundation model for zero-shot nuclei classification

Yinuo Xu, Jina Kim, Zijie Zhao, Zhi Huang

An ontology-aware vision–language foundation model for zero-shot nuclei classification across tissues and unseen cell classes.

  • MEDICAL IMAGE
  • VISION-LANGUAGE MODEL
  • DEEP LEARNING
Venue: MMFM-BIOMED @ CVPR 2025
Access: Not public
05

Recent activity

Updates

  1. Released the Piece of CAKE preprint on microsecond-scale adaptive execution. (opens in a new window)
  2. Survivorship Bias in Industrial Database Workloads received the CIDR 2026 Best Paper Award. (opens in a new window)
  3. CodeImprove appeared at the 47th IEEE/ACM International Conference on Software Engineering. (opens in a new window)
  4. Joined the Penn Database Group to work on adaptive and self-tuning systems. (opens in a new window)
06

Academic record

Education

  1. PhD, Computer and Information Science

    University of Pennsylvania

  2. MA, Statistics and Data Science

    The Wharton School University of Pennsylvania

  3. BE, Computer Science

    Tianjin University