Justin Zhao
I build LLM systems, agentic workflows, and AI products for real users and operational settings.
I have been developing softwares for 8 years and focusing on AI harness and infrastructure building for 3 years. I build production LLM systems, agentic orchestration frameworks, and applied AI research pipelines, with experience from founding a startup engineering team to shipping production LLM systems today. My products have earned hundreds of stars on Github and scaled to over millions user interactions across platforms, and I build tools that let both technical and non-technical users deploy production-grade AI systems. I architect full-stack agentic systems end to end, including a drag-and-drop LLM workflow builder with a FastAPI and Celery backend and an agentic trading framework with switchable model providers and multi-agent loops. I build robust agentic systems, applying systematic evaluation methodology, cross-validated benchmarking, and bias-aware metrics to production decisions and I've published peer-reviewed papers on top journals. I translate technical work across audiences, mentoring next generation engineers and scholars, briefing leadership, and representing my work externally. I'm looking for challenging problems in distributed agentic systems, production AI infrastructure, and robust agentic system at scale.
Timeline
From a price monitor script in 2019 to production LLM systems and AI products today, each phase added engineering depth, product experience, and operational rigor.
Democratizing LLM Development: NoCode Platform
Architected a production NoCode LLM application using React Flow for visual orchestration and a FastAPI/Celery backend for asynchronous multi-agent execution across multiple OpenRouter endpoints. Users build complex LLM pipelines through drag-and-drop, reducing weeks of onboarding to minutes of configuration. Triaged bugs reported by users and rebalanced LLM request scheduling against available resources to keep the product meeting real community needs. 300+ stars on GitHub.
Price Monitor 2.0: Agentic Trading System
Built an agentic trading framework where agents fetch live market data, build context autonomously, generate decisions, and review each other's outputs before acting. Customizable orchestration layers bring context-aware, auditable trading to non-technical users without engineering overhead. Coached non-technical users through platform mechanics and risk considerations during high-volatility moments as user enablement, and ran product demonstrations for prospective users and investors to translate technical capability into a credible narrative. Prioritized the roadmap by frequency and severity across the full user base rather than raw request volume, avoiding vocal minority bias where a small number of vocal users would have degraded the experience for the majority, and when declining a minority request explained the tradeoff directly to the affected users and offered an alternative workaround. 100+ stars on GitHub.
Research Journey
Ph.D. Research Assistant, National Science Foundation, University of Texas at Dallas. Developed an automated pipeline that processed 10GB+ of noisy corpora, reducing human annotation time by 95% and converting a 3-month manual project into 4 hours of parallelized GPU computation. Designed an evaluation framework to measure the LLM biases, providing guidance for other researchers to responsibly use LLMs. Partnered with multiple research groups to translate requirements into technical specifications, coached junior researchers on AI engineering best practices, and established working guidelines for future iterations for multiple projects.
M.S. Computer Science
University of Colorado Boulder training in machine learning and cloud computing, applied to the systems and products built above.
Research Assistant, University of Texas at Dallas
Ran quantitative and text-as-data pipelines for research groups, covering data collection, knowledge mining, visualization, and machine learning for text.
Data Analyst, ByteDance
Applied analytical rigor to advertising strategy across multiple product lines, consistently driving traffic growth through data-informed decisions. Brought an automation mindset to reporting and campaign analytics, shifting the team from manual data collection to strategic work.
Where It Began: Price Monitor, 2019
Built the first iteration of Price Monitor in 2019, a rule-based system for tracking prices and surfacing actionable signals. The same core problem that drove the 2019 version, making market information legible and actionable without constant manual monitoring, is what the 2026 agentic system solves at a fundamentally different level of autonomy.