Kevin Hernandez
Senior Data Engineer, AI & Data Solutions
Designing scalable data pipelines, lakehouse architectures, and real-time processing across AWS, GCP, and Azure. Expert in Apache Spark, Databricks, Delta Lake, Kafka, and AI/ML integration.
Scroll
— About
Senior Data Engineer with extensive experience designing scalable data pipelines, lakehouse architectures, and real-time processing across AWS, GCP, and Azure.
Led the delivery of a HIPAA-compliant data platform for 500+ clinicians, achieving 93% extraction accuracy and reducing manual effort, and built a high-throughput image generation pipeline supporting 5,000+ requests per month.
Expert in Apache Spark, Databricks, Delta Lake, Kafka, and AI/ML integration (RAG, LLMs), with strong data-modeling skills for analytics and BI.
My goal is to leverage this expertise to drive data-driven solutions that improve operational efficiency and support strategic decision-making.
— Experience
Key Responsibilities & Achievements
- Designed and built the backend data architecture for ZOYO, an AI powered real estate platform offering 8+ tools including virtual staging, interior/exterior design, day to night conversion, and image enhancement, scaling to 100+ monthly active users.
- Developed a modular NestJS backend with domain‑driven data models, REST and WebSocket APIs, and JWT authentication, delivering all platform services with 99.9% uptime on AWS ECS with auto‑scaling.
- Built a high throughput image generation and processing pipeline using RunPod and Flux models, coordinating data flow across storage, queues, and inference services to handle 5,000+ image requests per month at an average of 6 seconds per response.
- Engineered data integrations with GPT 5 and Claude Sonnet 4.5 to power property description generation and context aware design recommendations, improving user engagement metrics by ~38%.
- Implemented an Angular frontend with NgRx state management and RxJS reactive data streams for optimistic UI updates on image generation workflows, reducing perceived load times by ~45%.
- Integrated Stripe billing data and built an ROI calculator to support pricing decisions, contributing to a 22% increase in free to paid conversion within 3 months of launch.
Technologies Used
NestJSAWS ECSRunPodFluxGPT-5ClaudeAngularNgRxStripeWebSockets
Key Responsibilities & Achievements
- Led the design and launch of a HIPAA-compliant AWS data platform (EKS, Bedrock, SageMaker) with RBAC and audit logging, enabling real-time voice, text, and document data pipelines for 500+ clinicians and care coordinators in the first year.
- Engineered a low latency document processing pipeline combining AWS Textract with Amazon Bedrock and RAG based EHR context injection, sustaining over 93% extraction accuracy and increasing downstream clinical report completeness by ~28%.
- Built a lakehouse architecture using Databricks and Delta Lake to unify structured and unstructured clinical data, and applied Apache Spark (PySpark, Spark SQL) with star schema data modeling to process high volume datasets for downstream analytics and BI reporting.
- Built automated data orchestration systems with MCP‑compliant interfaces using n8n and GitHub Actions for care‑plan routing, escalation handling, and follow‑up automation, cutting manual data‑entry errors by 30% and improving clinical operational efficiency.
- Scaled FastAPI microservices on EKS with containerized inference workloads via Amazon ECR, enforcing HIPAA aligned network policies and PHI data isolation, achieving 99.8% uptime while reducing cloud spend by ~20% through autoscaling.
- Implemented continuous data quality and LLM evaluation pipelines with RAG relevance scoring, hallucination detection, and PHI redaction monitoring, enabling proactive regression detection across production environments.
Technologies Used
AWSEKSBedrockSageMakerDatabricksDelta LakeApache SparkTextractRAGMCPn8nFastAPI
Key Responsibilities & Achievements
- Containerized and deployed DeepHow's platform on GCP (GKE, Cloud Run, Cloud Storage) using Docker and GitHub Actions CI/CD pipelines, improving deployment reliability and reducing infrastructure costs by ~20%.
- Built internal data pipelines and star schema data models using Apache Spark (PySpark, Spark SQL) and FastAPI, enabling clients to track workforce knowledge usage through Next.js dashboards and cutting analytics turnaround from days to hours.
- Built Power BI dashboards and semantic models using DAX and Power Query/M to visualize workforce training and knowledge usage data, giving clients self-service analytics on top of the underlying data pipelines.
- Developed core FastAPI backend services and GraphQL APIs powering DeepHow's AI knowledge engine, improving knowledge transfer workflows by ~30% and reducing customer support tickets by ~18%.
- Built and maintained a customer-facing data application using Next.js with TypeScript and GraphQL, serving 50,000+ monthly users across manufacturing and industrial clients.
- Mentored junior engineers on data pipeline design, TypeScript, and Docker best practices, and partnered with product managers to improve sprint planning and reduce post-release defects by ~25%.
Technologies Used
Apache SparkPower BIGCPGKEFastAPIGraphQLNext.jsTypeScriptDockerGitHub Actions
Key Responsibilities & Achievements
- Designed and optimized PostgreSQL database schemas using conceptual, logical, and physical data modeling techniques, including star schema design, for IoT device and real estate data, improving query performance by ~35%.
- Integrated Kafka to stream real-time IoT sensor data from smart buildings, enabling live tracking of energy usage, occupancy, and facility operations while reducing data latency by ~40%.
- Built backend services and REST APIs using Django and Django REST Framework to support smart building and real estate portfolio management, accelerating feature delivery by ~25%.
- Developed APIs with Firebase authentication, validation, and error handling, improving API reliability and reducing error incidents by ~20%.
- Deployed backend services on Azure (App Service, Functions, Event Hubs) using Docker, and built Azure Event Hub based streaming pipelines to complement Kafka for real-time data ingestion, achieving 99.9% uptime and reducing infrastructure costs.
- Automated internal facility and portfolio reporting workflows using Power Automate and Power Apps with custom connectors, reducing manual effort for smart building operations teams.
Technologies Used
PostgreSQLKafkaData ModelingDjangoAzureEvent HubsPower AutomatePower AppsFirebaseDocker
Key Responsibilities & Achievements
- Built a full stack HR and recruiting platform, developing a React single-page application with MUI components on the frontend and FastAPI services backed by MongoDB on the backend, accelerating candidate processing and reducing onboarding time for outsourcing clients.
- Delivered end-to-end product features across the full stack, including candidate tracking, job posting management, and employee onboarding interfaces, reducing manual HR workload for clients.
- Participated in Agile ceremonies, code reviews, and CI workflows with Docker across frontend and backend codebases, improving sprint velocity by ~15%.
Technologies Used
ReactMUIFastAPIMongoDBDockerAgile
— Education
Achievements
- Specialized in Data Engineering and AI
- Core focus on Algorithms and Distributed Systems
Activities & Leadership
- AI Research Group Member
- Open Source Contributor
Relevant Coursework
Data Structures & AlgorithmsArtificial IntelligenceSoftware EngineeringDatabase SystemsCloud ComputingMachine Learning
— Skills
Technical Skills
Data Engineering & Pipelines
Apache Spark
Databricks
Delta Lake
Lakehouse
Kafka
Azure Event Hub
AWS Lambda
Step Functions
ETL/ELT
RAG Pipelines
Streaming
Data Modeling
Star Schema
Dimensional Modeling
Conceptual/Logical/Physical
Data Warehousing
Power Platform & BI
Power BI (DAX, Power Query/M)
Semantic Modeling
Power Apps
Power Automate
Custom Connectors
Cloud & DevOps
AWS
GCP
Azure
Docker
GitHub Actions
Jenkins
Vercel
CI/CD
Databases
PostgreSQL
MySQL
MongoDB
DynamoDB
Firebase
Supabase
Redis
Data Science & AI/ML
OpenAI
Claude
Gemini
DeepSeek
LangChain
LangGraph
LangSmith
RAG
MCP
Multimodal AI
Multi-Agent AI
Languages
Python
JavaScript (ES6+)
TypeScript
Go
Java
C#
Backend
Node.js
ExpressJS
NestJS
Django
Flask
FastAPI
REST
GraphQL
Laravel
Frontend
React
Vue.js
Next.js
Angular
Tailwind
MUI
SCSS
Workflow Automation
Vellum AI
Zapier
Make
n8n
Pipedream
AI Coding Assistants
Claude Code
Cursor AI
Google Antigravity
GitHub Copilot
AI No/Low-Code Platforms
Lovable
Bolt.new
v0
Replit Agent
Builder
Base44
Bubble
— Certifications
Professional Certifications
Industry-recognized credentials demonstrating expertise in AI, cloud computing, and modern technologies.
AI & Machine Learning
Generative AI Essentials on AWS
AIS Academy
AI & Machine Learning
AI in Finance Micro-Credential Certificate
IMA | Institute of Management Accountants
AI & Machine Learning
Generative & Agentic AI
IBM Consulting
AI & Machine Learning
Generative & Agentic AI Expert Architect
IBM Consulting
Cloud & DevOps
Azure Stack HCI Foundation Level
Dell & Microsoft
AI & Machine Learning
NABA Artificial Intelligence Certified
NABA Inc.
— Services
Project Catalog
Strategic consulting packages designed to accelerate your data platform, analytics, and AI initiatives with proven methodologies.
ETL/ELTArchitecture
Data Pipeline Architecture Audit
ETL/ELT Assessment & Modernization Roadmap
12 days
View details
DatabricksDelta Lake
Lakehouse Architecture Design
Databricks, Delta Lake & Spark Platform
14 days
View details
KafkaStreaming
Real-Time Streaming Pipeline
Kafka & Event-Driven Data Architecture
14 days
View details
Power BIDAX
Power BI Analytics Platform
Semantic Models, DAX & Self-Service BI
10 days
View details
Star SchemaData Modeling
Data Warehouse & Dimensional Modeling
Star Schema Design for Analytics at Scale
10 days
View details
RAGLLM
RAG Data Platform Blueprint
Retrieval Pipelines for LLM Applications
12 days
View details
LLMAPI
LLM Data Integration Strategy
GPT, Claude & Gemini in Your Data Stack
8 days
View details
MCPOrchestration
MCP Data Orchestration Plan
Model Context Protocol for Data & AI Workflows
12 days
View details
InferenceGPU
AI Media Processing Pipeline
High-Throughput Image & Inference Workloads
14 days
View details
— Testimonials
Client Reviews
Trusted by CTOs, Engineering Leaders, and Founders to deliver enterprise-grade data platforms and AI solutions
Mark Stephens
CRO
DeepHow
"Kevin transformed our analytics capability. The data pipelines and Power BI dashboards he built took our reporting turnaround from days to hours, and our clients finally got self-service insight into their workforce training data. His work directly strengthened our customer retention."
Project: Analytics Data Pipelines
10+
Years Experience
500+
Clinicians on Platforms Built
99.9%
Production Uptime
93%+
RAG Extraction Accuracy
— Contact
Let's work together
If you would like to discuss a project or just say hi, I'm always down to chat.