Vicino AI, building image, video, and 3D generation workflows.
Backend and Applied AI Engineer
Xing Gao
Backend systems for AI products, research platforms, and distributed workflows.
I work on the backend pieces behind AI products.
I'm a backend and applied AI engineer with 3+ years building production services, distributed systems, and LLM-driven workflows. Most of my work sits behind the UI: durable jobs, async workers, retries, provider routing, evaluation loops, tracing, and reliability.
PayPal backend services for compliance and sanctions workflows.
Research assistant work on AI education and knowledge graph systems.
Backend and AI systems shipped in production.
2026 - Present
Software Engineer, Applied AI
Vicino AI
Productionized multimodal AI generation services for image, video, and 3D workflows using Python, FastAPI, Redis Streams, SQL-backed job state, async workers, retries, status updates, monitoring, and AWS deployment automation.
- Image, video, 3D
- Durable async jobs
- Evaluation loops
2021 - 2023
Software Engineer
PayPal
Maintained Java and Spring Boot backend services for compliance and sanctions workflows, supporting transaction screening, account review, regulatory workflows, and policy enforcement.
- 5+ OTEL migrations
- 200+ markets
- SignalFx and Splunk
2025
Research Assistant, AI Education Platform
UIUC AI Education Platform
Built research platform infrastructure across SQL assessment, automated feedback, knowledge graph APIs, LLM-assisted concept extraction, human-reviewed relationship editing, instructor analytics, and research publication artifacts.
- Neo4j knowledge graph
- FastAPI and React
- 9k+ line contribution
Earlier
Backend Engineering Internships
Bili and NaviData AI
Built LangGraph-backed Google Docs/Sheets workflows with diff/patch editing, schema validation, OAuth2, retries, and audit logs. Also developed a FastAPI, Azure SQL, and Redis recommendation backend, reducing P95 latency from 800ms to 200ms.
- LangGraph
- FastAPI
- Azure SQL and Redis
Selected systems, research tools, and product prototypes.
UIUC Research Assistant
AI Research Platform
Tools
Built platform pieces for an AI-assisted engineering education system: SQL assessment, automated feedback, student submission tracking, course-material ingestion, knowledge graph construction, human-reviewed concept relationships, instructor analytics, and publication analysis workflows. The work supported analysis of 3,187 interaction turns across 756 students.
Developer tooling
TokenCause
CodeTools
Built a local-first observability CLI for AI coding sessions. TokenCause reads session traces, detects retry loops, broad exploration, repeated artifacts, long tool outputs, context drift, and cache-heavy billing patterns, then renders diagnosis-first HTML reports, dashboards, and JSON output for automation.
Cozad startup project
Moxie
Tools
Moxie turns course notes, slides, or PDFs into short playable games. I built the MVP flow around AI authoring jobs, creator approval, game publishing, play URLs, QR sharing, and analytics for choices, endings, drag attempts, and drop-off points.
CacheHacks software track
Fluxa
DemoTools
Fluxa is an AI-assisted procurement war room for hardware teams. It parses editable BOM data, connects external supply signals to affected parts, scores risk with deterministic math, generates an analyst brief, compares what-if actions, and creates an operator-approved execution pack.
Leader
Follower
Follower
Log
Heartbeat
Recovery
Distributed systems coursework
Distributed Systems Suite
Tools
Built a distributed file system with Raft consensus, leader election, heartbeat failure detection, log replication, and network partition handling. The system ran across a 10-node cluster and validated recovery behavior under node failures with 2-fault tolerance.
Smaller builds that round out the story.
Research agent
Repo
DeepSeeker
Self-healing deep research agent with planner, web workers, critic-triggered retries, cited reporting, persisted state, and trace-based debugging.
Agent workflow
Google Docs Integration
LangGraph document generator with parallel section generation, Google Docs creation, OAuth fallback, image/table insertion, and section-by-section editing.
LLM simulation
Simulation Agent
Mesa-based simulation runtime where LLM agents make natural-language decisions, collect behavior data, and generate analysis reports.
A practical stack for backend systems, AI workflows, and production observability.
Languages
Backend services, systems coursework, AI tooling, and frontend integration.
Backend Platforms
Service boundaries, APIs, background jobs, integration pipelines, and reliability work.
Data Systems
Relational stores, caches, graph modeling, semantic retrieval, and analytics storage.
AI Workflows
LLM applications that need orchestration, evaluation, retries, and human-readable outputs.
Cloud & DevOps
Deployment, runtime monitoring, service diagnostics, and developer workflow tooling.
Reliability
Tracing, metrics, service health, distributed-system behavior, and incident debugging.
Resume
Backend, AI infrastructure, and applied AI systems.
A compact version of my experience, projects, and technical stack for recruiter review.
Social Profiles
Email Me
happyxgao@gmail.com
Open For
Backend, AI infrastructure, and applied AI roles.