Hongyu Li

Research Interests

Education

Skills

• –Tech lead manager of a 5-engineer team, setting technical direction, prioritizing reliability work, and coordinating execution across engineering and product stakeholders.
• –Led ERP platform quality and stability initiatives across legacy integration systems with significant stability and maintainability challenges, improving master data pull reliability and enterprise customer workflows.
• –Drove ERP on App Studio enablement, migrating ERP integrations onto the App Studio platform to unlock support for a broader range of ERP systems and expand the addressable customer base.

• –Tech lead for a 6-engineer team building App Studio from 0 to production, a no-code integration platform enabling organizations to build custom procurement and AI-agent workflows in days.
• –Designed workflow primitives for triggers, actions, branching logic, and agent-orchestrated system interactions.
• –Built platform infrastructure including a Connector SDK, execution engine, and job scheduler framework supporting dynamic job creation with flexible scheduling patterns, adopted by multiple product teams.
• –Led product infrastructure work, driving a major database migration project to improve scalability and data model flexibility.
• –Extended the platform to support AI agent use cases, treating agents as first-class entities that orchestrate workflows across multiple systems and enable agentic procurement automation without custom code.

• –Tech lead for a 5-engineer core payments team owning Pay-in and Pay-out APIs for enterprise merchants.
• –Architected the Pay-in service from 0 to production, scaling to 180K transactions per day across 5 countries with 20+ payment methods and generating millions in annual revenue.
• –Built core payment infrastructure including merchant configuration, vendor gateway, and routing engine.
• –Led a large-scale database migration to support horizontal scalability and long-term growth.
• –Worked with Java, Spring Boot, MySQL, and AWS.

• –Led Code AI infrastructure projects, building a unified indexing and training pipeline, standardized service framework, and shared inference infrastructure across Neural Code Search, Smart Code Completion, Getafix, and Aroma.
• –Built Smart Code Completion from 0 to 1 with 200K adoptions per day and Neural Code Search with 100K+ queries per day.
• –Published research on code search and ranking at ICSE, FSE, and MAPL, and presented at FSE 2019.

• –Built the NUX tip framework for Instagram Business Insights
• –Frontend NUX tip dynamic rendering on iOS

• –Built UI testing automation for Chromecast Android app

Open-domain question answering system with Freebase

• –Built an end-to-end QA system with Freebase from scratch
• –Built language models on matched relations using search engine retrieval

Web app teaching hand gestures with Leap Motion

• –2nd Place at SteelHacks 2015

Original side-scrolling game written in Python