M.S. Software Engineering Systems student at Northeastern University, seeking software engineering jobs in full-stack development and AI applications.
I build full-stack products and AI agent systems with React, TypeScript, Java, Spring Boot, PostgreSQL, LangGraph, and modern cloud/database tooling.
An AI-powered AIOps platform that automates incident diagnosis through multi-agent collaboration and turns resolutions into reusable runbooks.
- Built a polyglot microservice system with Spring Boot, FastAPI, React, Redis Streams, pgvector, and LangGraph.
- Implemented an incident-response pipeline for triage, evidence collection, root-cause diagnosis, remediation planning, and runbook generation.
- Designed live-streamed reasoning so on-call engineers can follow agent decisions in real time.
- Used Redis Streams to support durable event-driven workflow updates across backend services.
- Integrated pgvector-based semantic retrieval to help agents reuse historical incident knowledge.
- Frontend: React 18, TypeScript
- Backend: Spring Boot 3, FastAPI
- AI / Agent Framework: LangGraph
- Database: PostgreSQL, pgvector
- Messaging / Streaming: Redis Streams
- Deployment: Docker Compose
- Architecture: Microservices, Multi-Agent System, Event-driven Architecture, RAG
- GitHub: RunbookAgent
An automated multi-step AI career agent that helps users upload resumes, configure job preferences, search for matching opportunities, score job compatibility, and generate personalized cover letters.
- Built an interactive full-stack dashboard using React and TypeScript, allowing users to upload PDF resumes, configure job preferences, and monitor autonomous job-matching tasks in real time.
- Used the OpenAI GPT-5 API and LangGraph to orchestrate stateful AI workflows for resume parsing, LinkedIn search, compatibility scoring, and personalized cover letter generation.
- Developed an asynchronous FastAPI backend integrated with the LinkedIn API and APScheduler to support scheduled job-matching workflows.
- Implemented automated daily email notifications using SendGrid to deliver personalized job recommendations to users.
- Employed PostgreSQL with pgvector to manage a RAG pipeline for resume and job matching, enabling semantic comparison between user profiles and job descriptions.
- Designed an adaptive threshold algorithm that dynamically adjusts match-score requirements to ensure users receive 10 personalized job recommendations per day.
- Deployed containerized services with Docker to Railway, exposing REST APIs and async task queuing to support concurrent multi-user agent workflows.
- Frontend: React, TypeScript
- Backend: FastAPI, Python
- AI / Agent Framework: OpenAI GPT-5 API, LangGraph
- Database: PostgreSQL, pgvector
- Scheduling / Automation: APScheduler, SendGrid
- External API: LinkedIn API
- Deployment: Docker, Railway
- Architecture: Full-stack Application, RAG Pipeline, Async Task Queue, Agent Workflow
- GitHub: Job Matching AI
- Live Demo: welcoming-mindfulness-production-e40c.up.railway.app
A full-stack food ordering application with secure authentication, restaurant menu browsing, order management, PostgreSQL persistence, and AWS deployment.
- Built RESTful backend APIs using Spring Boot to support user authentication, menu browsing, cart management, and order processing.
- Developed a React frontend for customers to browse food items, place orders, and interact with the application through a responsive user interface.
- Integrated PostgreSQL for persistent storage of users, orders, menu items, and application data.
- Implemented Spring Security authentication and authorization to protect user sessions and backend resources.
- Deployed the application using AWS services including RDS, ECR, and App Runner.
- Frontend: React
- Backend: Spring Boot, Java
- Security: Spring Security
- Database: PostgreSQL, AWS RDS
- Deployment: AWS ECR, AWS App Runner
- Architecture: Full-stack Web Application, REST API
- GitHub: Online Order
An Android music streaming client built with Kotlin, Jetpack Compose, MVVM, and the Repository Pattern. It supports playlist browsing, playback controls, favorites, and offline persistence.
- Built a native Android music streaming client using Kotlin and Jetpack Compose.
- Implemented playlist browsing, music playback controls, favorites, and offline persistence.
- Used MVVM architecture and the Repository Pattern to separate UI, business logic, and data access.
- Integrated Retrofit and OkHttp for network communication with remote music APIs.
- Used Room database for local persistence and offline data storage.
- Applied Hilt for dependency injection and Coil for efficient image loading.
- Mobile: Kotlin, Jetpack Compose
- Architecture: MVVM, Repository Pattern
- Networking: Retrofit, OkHttp
- Database: Room
- Dependency Injection: Hilt
- Image Loading: Coil
- Architecture Type: Native Android Application
- GitHub: miniSpotify
Java, TypeScript, JavaScript, Go, SQL
React, React Native, Expo, Jetpack Compose
Spring Boot, Spring Security, NestJS, FastAPI, Node.js
PostgreSQL, MongoDB, Redis, Elasticsearch, pgvector
LangGraph, LangChain, RAG, OpenAI API
Docker, AWS, GCP, GitHub Actions
- Portfolio: jiajun-portfolio.vercel.app
- LinkedIn: linkedin.com/in/jiajun-zheng1012
- GitHub: github.com/jjMurphy1012
- Email: zheng.jiaju@northeastern.edu