AI APPLICATIONS / BACKEND ENGINEERING

Evan Chen

AI Application Development / Backend Engineering

Focused on traceable, testable, and production-oriented Agent and RAG systems.

OPEN TO WORK

AI Application / Backend Engineering Internships

M.S. student · Available to connect · Building in public

I turn Agents, RAG, and backend systems into products you can verify.

Technical Strengths

Agents: Agent Loops, tool calling, context management, planning and state progression, permissions, recovery, and validation

RAG: ingestion, structured chunking, embedding + BM25 hybrid retrieval, query rewriting, reranking, context engineering, citations, and evaluation

AI coding: Claude Code / Codex for code understanding, feature development, debugging, tests, refactoring, and result verification

Agent systems: using OpenClaw / Hermes with an understanding of tool scheduling, context passing, execution loops, state management, and result write-back

Backend engineering: Java application layers, Redis caching, MQ-based asynchronous processing and decoupling, plus FastAPI / React development

View projects→ Read technical notes→

CURRENT FOCUS

Agent Runtime · RAG Evaluation

KNOWLEDGE SERIES

Blog Publishing Automation→Claude Code Agent Runtime→RAG Evidence Engineering→NexaRAG Workflow Notes→RunLens Agent Runtime→

2 complete AI projects

56 technical notes

5 knowledge series

AI Projects

Agent / RAG Projects

Projects around AI application development, tool calling, hybrid retrieval, runtime tracing, and backend engineering.

AI Application Development / Backend Engineering

2026

NexaRAG

A RAG question-answering system for 3C digital product support, built around an evidence chain from ingestion and hybrid retrieval to context assembly, answer verification, and evaluation.

RAGAgentFastAPILangGraphpgvector

View project ->

Agent Runtime / AI Application Engineering

2025

RunLens

A verifiable, observable, and plugin-extensible local Agent Runtime that connects model loops, tool permissions, recovery, artifact validation, and runtime evidence in one real execution chain.

Agent RuntimeTool CallingFastAPIReactObservability

View project ->

Notes