Available for senior IC & staff-level opportunities

Xiaohui Chen

I build

Building intelligent, high-scale product infrastructure at DoorDash and Amazon for 10+ years.

  • Sugar Land, TX
  • xiao.chen.hui@outlook.com
  • (281) 299-0865
  • LinkedIn
View ProjectsDownload Resume

Career Highlights

10+ years building product infrastructure at companies serving hundreds of millions of users.

DoorDash

Senior Software Engineer

Sunnyvale, CA

Nov 2022 – Present

DashPass Subscription Platform

15M+
Subscribers

Architect the core DashPass platform across eligibility, billing, AI-powered reorder, and real-time event streaming. Drove a $40M+ ARR lift via a billing-state machine that cut involuntary churn by 20%.

  • Led the LangGraph + Claude reorder agent now serving millions of personalized recommendations per day
  • Owned p99 latency cuts from 180ms → 60ms on the eligibility gRPC service
  • Shipped DashPass for Teams (B2B) end-to-end across web, GraphQL, and Stripe billing
  • TypeScript
  • Python
  • Go
  • React
  • GraphQL
  • gRPC
  • Kafka
  • AWS

Amazon

Software Development Engineer II

Seattle, WA

Aug 2016 – Nov 2022

Amazon Prime · Sponsored Products

200M+
Prime Members

Scaled the Prime membership conversion funnel and the Sponsored Products ad retrieval pipeline. Built ML feature platforms supporting 10B+ daily impressions across the Amazon retail ecosystem.

  • Owned ad retrieval ranking with sub-100ms p99 across Prime Day traffic peaks
  • Built an LLM-based ad-copy quality classifier that cut manual review queue by 30%
  • Tech-led the Prime signup A/B framework that drove a measurable lift in conversion
  • Java
  • Node.js
  • Python
  • React
  • DynamoDB
  • Kafka
  • Elasticsearch
  • AWS

Featured Projects

Production systems built at scale.

DashPass Eligibility Microservice mockup
DoorDash

DashPass Eligibility Microservice

Architected a gRPC eligibility service with PostgreSQL + Redis caching, powering real-time access gates across the checkout and Teams portal for 15M+ subscribers.

↓ p99 latency 180ms → 60ms
  • gRPC
  • PostgreSQL
  • Redis
  • TypeScript
  • AWS
AI-Powered Reorder Engine mockup
DoorDash

AI-Powered Reorder Engine

Built a LangGraph two-agent pipeline with pgvector retrieval and Anthropic Claude via MCP, surfacing personalized reorder recommendations with natural-language rationales on the React home feed.

↑ Measurable reorder engagement lift
  • LangGraph
  • pgvector
  • MCP
  • AWS Bedrock
  • React
Billing State Machine & Churn Recovery mockup
DoorDash

Billing State Machine & Churn Recovery

Engineered a Go billing backend with Stripe payment intents, a grace-period state machine, and an automated retry scheduler, synchronized with the real-time member-state GraphQL layer.

↓ Involuntary churn by ~20%
  • Go
  • Stripe
  • Kafka
  • PostgreSQL
  • gRPC
DashPass for Teams Portal mockup
DoorDash

DashPass for Teams Portal

Shipped the B2B admin portal in Next.js SSR and the member home feed in React/TypeScript, backed by a NestJS provisioning API with Stripe B2B invoicing.

Serving 15M+ DashPass subscribers
  • Next.js
  • React
  • NestJS
  • TypeScript
  • Stripe
Sponsored Products Ad Retrieval mockup
Amazon

Sponsored Products Ad Retrieval

Scaled the Java Spring Boot ad retrieval pipeline with Elasticsearch multi-stage ranking (candidate generation, feature reranking, and SageMaker neural ranking) under Prime Day peak traffic.

< 100ms p99 under Prime Day load
  • Java
  • Spring Boot
  • Elasticsearch
  • SageMaker
  • Kafka
LLM Ad Copy Quality Classifier mockup
Amazon

LLM Ad Copy Quality Classifier

Integrated an OpenAI API classifier into the Java ad ingestion pipeline via a Python sidecar, validated with a custom LLM eval harness, precision/recall baselines, and 2 weeks of shadow-mode testing.

↓ Manual review queue by ~30%
  • OpenAI API
  • Python
  • Java
  • gRPC
  • LangSmith

  • 15M+

    DashPass Subscribers Served

  • 200M+

    Amazon Prime Members Reached

  • ~20%

    Involuntary Churn Reduction

  • 35%

    Post-Launch Incident Reduction

Technical Skills

Polyglot stack across product, infra, and AI.

Languages
  • TypeScript
  • Python
  • JavaScript
  • Go
  • Java
  • SQL
Frontend
  • React
  • Next.js (SSR/ISR)
  • Apollo GraphQL
  • Redux Toolkit
  • Tailwind CSS
  • D3.js
Backend
  • Node.js
  • NestJS
  • FastAPI
  • Spring Boot
  • Express
  • gRPC
  • GraphQL
  • REST
AI / ML
  • LangGraph
  • LangChain
  • RAG Pipelines
  • MCP
  • AWS Bedrock
  • OpenAI API
  • pgvector
  • LangSmith
  • RAGAS
Data & Streaming
  • Apache Kafka
  • Flink
  • PostgreSQL
  • DynamoDB
  • Redis
  • Elasticsearch
  • Snowflake
Cloud & Infra
  • AWS (EKS, Lambda, S3, RDS, SageMaker)
  • Docker
  • Kubernetes
  • Helm
  • Terraform
  • GitHub Actions
Observability
  • OpenTelemetry
  • Prometheus
  • Grafana
  • Datadog
  • Distributed Tracing
  • SLO Management

Certifications

AI Engineer Agentic Track: The Complete Agent and MCP Course

Udemy · Jan 2026

ID: UC-AI-AG-2026-XC

Adding Knowledge to LLMs

NVIDIA · Nov 2025

ID: NV-LLM-2025-XC

Education

  1. M.S. Computer Science

    The University of Texas at Austin

    2015 – 2016

  2. B.S. Computer Science

    The University of Texas at Austin

    2011 – 2015

Built with v0