Yi Shen

Senior Software Development Engineer & Architect

Building Tier-1 real-time ML platforms handling petabytes of data and millions of QPS at Amazon. Homelab enthusiast running Kubernetes clusters, smart home automation, and enterprise-grade infrastructure.

Download Resume Contact Me View Homelab

About Me

Senior Software Development Engineer with 8 years at Amazon, architecting Tier-1 real-time ML platforms handling petabytes of data & millions of QPS with 99.99% availability. Organizational leader driving cross-company innovation by influencing strategic technology roadmaps to unlock new capabilities and 60% cost reductions.

Passionate about AI automation, distributed systems, and delivering high-impact solutions with strong ownership and engineering excellence. Outside of work, I run an extensive homelab featuring Kubernetes clusters, NVR systems, and smart home automation.

Work Experience

Senior SDE (SDE 3)

Amazon.com Services LLC

July 2023 - Present
  • Architect of a Tier-1 real-time self-service ML feature platform serving hundreds of internal scientists, handling petabyte-scale data and millions of QPS with 7ms latency and 99.99% availability.
  • Designed and led implementation of a centralized Serverless Control Plane (AWS API Gateway, Lambda, Step Functions), reducing infrastructure provisioning time by 90%.
  • Led strategic evaluation of distributed databases (TiDB, Aerospike, FoundationDB) and emerging compute frameworks (Ray, Pulsar) via deep-dive POCs and cross-company collaboration, achieving 60% cost savings.
  • Developed an internal AI assistant using RAG, MCP and Strands Agent, reducing scientist onboarding from 2 weeks to less than 2 days.
  • Optimized platform retrieval performance, reducing p99 latency from 900ms to 40ms, saving Amazon $50M+ in fraud losses annually.
  • Served as a Security Reviewer for cross-team initiatives, auditing designs for Kubernetes clusters and AWS infrastructure.

SDE 2

Amazon.com Services LLC

March 2020 - June 2023
  • Implemented real-time automatic hotkey and anomaly detection to defend against bot attacks.
  • Pioneered automatic cross-account Infrastructure as Code (IaC) provisioning systems, reducing manual onboarding work by 90%.
  • Reduced metadata load latency from 30s to 4s through cache optimizations.
  • Engineered robust streaming data pipelines with AWS Kinesis, enabling real-time fraud detection with second-level latency.

SDE 1

Amazon.com Services LLC

June 2018 - March 2020
  • Contributed to building and maintaining the internal model training and inference platform for fraud detection, serving hundreds of internal scientists.
  • Developed a critical feature simulation framework that reduced iteration time from weeks to days by enabling offline simulation without impacting production.

Homelab & Projects

Enterprise-grade infrastructure running at home

☸️

Kubernetes Cluster

Highly available K8s cluster on Proxmox with 3 masters + 3 workers. Infrastructure as Code with Terraform, Ceph for distributed storage.

Kubernetes Proxmox Terraform Ceph
📹

Frigate NVR

AI-powered network video recorder with object detection. Hardware acceleration via Coral TPU and NVIDIA RTX 3070.

Frigate Coral TPU NVIDIA GPU AI
🏠

Home Assistant

Complete smart home automation hub integrating lights, sensors, cameras, and climate control with custom automations.

Home Assistant Zigbee Z-Wave MQTT
💾

TrueNAS Storage

Enterprise-grade NAS with ZFS for data integrity. 10GbE networking with VLAN segmentation for performance and security.

TrueNAS ZFS 10GbE iSCSI
🔒

OPNsense Firewall

Network edge security with multi-VLAN routing, WoL support, and IDS/IPS. Migrated from pfSense for improved features.

OPNsense VLANs Firewall IDS/IPS
🌐

Secure Access

Nginx Proxy Manager for internal service routing with SSL termination. Cloudflared tunnels for secure external access without exposing ports.

Nginx Proxy Manager Cloudflared SSL Zero Trust

Professional Skills

Languages

Java, Python, TypeScript, SQL, Shell

Cloud & Infra

AWS, Kubernetes, Docker, Flink, Git

AI & Data

LLM, RAG, MCP, Spark, Hadoop, EMR

System Architecture

Serverless, Microservices, Event-Driven, Caching

Education

University of Wisconsin-Madison

B.S. Computer Engineering, double major in B.S. Computer Sciences

GPA: 3.94 / 4.0 May 2018

Stanford University (Summer Session)

Intensive Studies In Computer Science Certificate

GPA: 4.2 / 4.3 Summer 2016