Funi Chen
Sr. Data Engineer | MD | AI Systems Architect | Full-Stack Builder
Bridging Clinical Reality and AI Engineering. Chat with my AI to see how I build secure, scalable systems.
About Me
The AI-Augmented Architect
I bridge the gap between Clinical Reality and AI Engineering.
My career is defined by a rigorous, first-principles approach: transitioning from Medical Doctor (M.D.) to Sr. Data Engineer at Amazon Pharmacy and Providence Health. I specialize in architecting the data infrastructure that makes healthcare AI possible—transforming messy clinical records into secure, model-ready assets.
Engineering Philosophy
Beyond enterprise platforms, I operate as a hands-on builder. I release independent apps under my personal developer alias, MangoMelon, utilizing an AI-First engineering workflow ("Vibe Coding") to rapidly prototype and ship production-grade software.
My current technical focus includes architecting Secure Agentic Systems (FastAPI/Gemini), developing Native Mobile Apps (Flutter), and implementing DevOps Pipelines. I don't just manage engineering teams; I build the systems they run on.
Professional Experience
Sr. Data Engineer (Official Title: Sr. BIE)
Amazon Pharmacy- AI-Driven Interfaces: Led the implementation of a Natural Language Querying (Text-to-SQL) capability, enabling 50+ non-technical stakeholders to query pharmacy data instantly, reducing ad-hoc ticket volume by 40%.
- Clinical Data Modeling: Engineered a Unified Clinical Data Model (SSOT) for Hub & Prescriber operations. Consolidated siloed datasets into a high-performance schema that serves as the foundation for downstream Machine Learning initiatives.
- Automated Governance: Built a config-driven Data Quality Framework using Python & AWS Glue. Replaced manual Excel validation with automated anomaly detection pipelines, ensuring 100% regulatory compliance.
Engineering Manager, Analytics
Providence Health- Technical Leadership & Team Transformation: Led a cross-functional squad of 5 (Architects, Devs, BIEs). Professionalized the engineering culture by implementing Agile/Scrum workflows and CI/CD standards, resulting in a 30% increase in delivery velocity.
- Architected Value-Based Care Platform: Directed the system design for a scalable financial analytics platform covering Profit & Loss (P&L). Unified complex clinical and financial data models to support strategic decision-making for executive leadership.
- DevOps & Code Governance: Established rigorous development environments and code review protocols (Git). Optimized query performance on Snowflake/Azure, ensuring high availability and low latency for mission-critical financial reporting.
Sr. Data Engineer, Tech Lead (Official Title: Sr. BIE)
Amazon- Scalable Healthcare Infrastructure: Architected a patient-centric Data Lakehouse for Quality Analytics. Optimized storage schemas to support real-time population health analysis, resulting in a 40% increase in decision-making velocity.
- Pipeline Automation: Engineered event-driven ETL workflows for global benefits utilization, replacing manual reporting cycles with automated pipelines. Achieved a 65% increase in operational efficiency by reducing data latency.
Principal Data Engineer (Clinical ML)
Providence Health- Clinical ML Pipelines: Productionized data workflows for predictive risk modeling (e.g., Chemotherapy outcomes). Translated complex medical protocols into feature engineering logic for automated risk stratification.
- Cloud Architecture Migration: Led the migration of legacy clinical data for 51 hospitals into a modern Snowflake/Azure architecture. Optimized ETL logic to improve query performance by 75% for critical patient safety reporting.
- Healthcare Data Standards: Designed scalable data marts compliant with HL7/FHIR standards, supporting real-time decision making for Population Health initiatives.
Engineering Ventures
SquishyDuck - Mobile Physics Engine
View on App StoreA high-performance stress relief app published on iOS. Features custom gesture recognition, vertex-based mesh deformation, and an optimized 60fps rendering loop using pure Dart.
Insight Agent Connect (Backend)
A secure multi-agent orchestration system. Designed with a hybrid architecture (FastAPI + Streamlit) featuring RAG pipelines, credential rotation policies, and strict access controls.
Portfolio AI Integration (DevOps)
A modular AI widget with a secure CI/CD pipeline. Implemented GitHub Actions for build-time secret injection to protect API keys in client-side code.