I am a software engineer focused on AI systems, agent workflows, and backend infrastructure.
My recent work centers on:
- building RAG and document-processing pipelines
- designing tool-integrated agent systems
- creating evaluation loops and workflow automation
- shipping multi-service applications with strong engineering guardrails
- combining backend reliability with product-minded delivery
I care about systems that are not only intelligent, but also observable, testable, and production-ready.
Resume Link: https://www.dropbox.com/scl/fi/489x42fcmbx9x6pkyvpbd/Hongxi_Chen_resume.pdf?rlkey=66bw3etqj1ikgtld9ptfddzkv&st=ie3ym350&dl=0
AI Systems
├─ RAG pipelines
├─ tool orchestration
├─ agent workflows
└─ evaluation and feedback loops
Backend Engineering
├─ distributed services
├─ API and service boundaries
├─ queue and event driven workflows
└─ observability and operational reliability
Engineering Style
├─ repository-first specs
├─ explicit workflow contracts
├─ deterministic guardrails
└─ human-controlled critical logic
|
AI-assisted investment research and risk platform built as a TypeScript monorepo. Includes a NestJS API, Next.js frontend, agent orchestration, trading integrations, document ingestion, and RAG workflows. Highlights: multi-service architecture, RAG pipeline, shared packages, PostgreSQL, Redis, observability, production-oriented backend design. |
|
|
Research-heavy Battleship project spanning game logic, web gameplay, and learning-based decision systems. Combines Java and web services with reinforcement learning based action selection. Highlights: PPO-based opponent service, Java to Python bridge, Spring backend, web integration, model serving, game AI under partial observability. Status: private research project |
Distributed e-commerce and logistics simulation projects focused on order lifecycle, delivery coordination, protocol handling, and robust backend interaction. Highlights: network communication, backend workflow design, simulated logistics, concurrent processing, systems-oriented engineering. |
|
Experimental database-oriented project inspired by Amazon-style catalog workflows, focused on schema design, query logic, and product data handling. Highlights: SQL thinking, data modeling, backend data flow, product catalog management. |
Together, these projects show a profile centered on applied AI systems, backend depth, workflow automation, and production-minded engineering. The goal is not just to build demos, but to ship systems with clear boundaries, measurable behavior, and practical value. |
1. FinSentinel
2. career-ops
3. mini-amazon-ups
4. mini-ups
5. amazon-database
|
I like building systems where LLMs are part of a larger product architecture instead of a standalone demo. |
I care about validation layers, explicit contracts, instrumentation, and feedback loops that make AI systems safer to operate. |
I enjoy the engineering side of services, APIs, databases, queuing, deployment, and the discipline required to keep systems maintainable. |
1. Build around clear interfaces
2. Keep critical logic deterministic
3. Instrument everything important
4. Use feedback loops to improve systems
5. Treat AI as part of an engineered workflow, not magic
- AI infrastructure
- applied agent systems
- RAG and retrieval pipelines
- backend and platform engineering
- developer tooling
- evaluation, observability, and reliability for AI products
