Hang (Eric) Zheng

2025-09 - Present

• Designed and developed Morfeus, an AI/LLM-powered market intelligence platform for FX and macro trading, spanning conversational research, War Room risk monitoring, data feeds, session recaps, and Trade Lab workflows.
• Architected a multi-agent orchestrator that aggregated and validated outputs from 33 agents, producing consensus-backed views on currency pairs, macro events, and trade ideas.
• Integrated Polygon, Event Registry, Myfxbook, COT, and structured internal research datasets to power agent reasoning, market regime classification, news sentiment analysis, and research workflows.
• Shipped production infrastructure for authentication, subscriptions, background jobs, PDF reporting, automated tests, and CI/CD across a Next.js/FastAPI/Postgres/Redis stack.

2020-08 - 2024-12

• Led the design, development, and maintenance of the LiveMe risk control platform using Golang, Kafka, and Elasticsearch.
• Refactored core modules from Java to Go, improving task scheduling efficiency and reducing resource consumption.
• Introduced Kafka-based real-time event streaming and Elasticsearch user profiling, reducing risk decision latency to seconds and cutting server costs by 30%.
• Built and maintained a data warehouse on AWS Redshift, integrating multiple risk sources for unified analytics and reporting.

2017-02 - 2020-08

• Contributed to the development and maintenance of the credit wallet backend system using Spring Boot, MyBatis, Hibernate, MySQL, Redis, and RabbitMQ, supporting over 100K DAU and peaking at 1M daily.
• Led the redesign of loan fund management and repayment modules, implementing circuit breakers, automated allocations, and streamlining third-party integration strategies for financing partners.
• Simplified integration processes for new lenders, reducing onboarding time and enhancing system maintainability.
• Received the Annual Best Employee award and monthly Culture Star award for outstanding technical leadership and team collaboration.

2016-05 - 2017-02

• Led development of user, product, and session management modules during a full rewrite of a legacy cloud gaming business system.
• Helped deploy a nationwide system with daily page views over 100 million and reduced database concurrent reads from over 500/s to approximately 30/s on non-peak days.
• Built a Flume to Kafka to JStorm to MongoDB/HDFS log analysis pipeline processing over 1 TB per day.
• Resolved high CPU usage and thread-safety issues in KafkaSpout and fixed offset skew; the system ran stably for more than 6 months.

2015-06 - 2016-05

• Central Hospital Appointment and Call Platform.