Payments and distributed systems
Designed a Go and Postgres limits platform for ACH and wire payments, including concurrency control with advisory locks and config-driven decisioning.
SENIOR SOFTWARE ENGINEER · BACKEND · FULL STACK · CLOUD
I build dependable systems for high-stakes workflows.
I’m Joseph Hernandez. For nearly six years, I’ve built and operated production software across banking, payments, data, and cloud infrastructure—from Go, Python, and Java services backed by Postgres to durable workflows and React applications.
Backend systemsGo, Java, Python, Postgres, DynamoDB
Frontend systemsTypeScript, JavaScript, React, Next.js, Tailwind CSS, micro frontends
Workflow designRestate, XState, DMN, event-driven architecture
Production ownershipAWS, Docker, OpenTelemetry, CI/CD, on-call
EXPERIENCE
Backend depth, with full-stack range.
My core work is designing resilient services and the workflows around them. I also lead user-facing delivery when the product needs to span the stack.
Durable orchestration
Built long-running workflow orchestration with Restate and XState to handle distributed transactions, retries, and stateful business processes.
Cloud delivery and operations
Ship AWS workloads across ECS, Fargate, Lambda, Kinesis, EMR, and S3, with Docker, CI/CD, OpenTelemetry tracing, and regular production support.
Frontend and technical leadership
Lead React and TypeScript application work, contribute across Go microservices, design migrations, and mentor early-career engineers.
Languages:Go · Java · Python · TypeScript · Scala
Backend systems:PostgreSQL · DynamoDB · Spring Boot · Restate · XState · DMN · Distributed locking and transactions
Cloud and delivery:AWS · ECS · Lambda · Fargate · Kinesis · Step Functions · Glue · S3 · Docker · Terraform · CI/CD · OpenTelemetry
Data engineering:Spark · Hadoop · EMR · Snowflake · Salesforce · Data lakes
Frontend and AI tooling:TypeScript · JavaScript · React · Next.js · Tailwind CSS · MFE architecture · Claude Code · Windsurf · Cursor · Codex · GitHub Copilot
SELECTED WORK
Projects you can inspect and try.
REAL-TIME AI · WEBSOCKETS · STATE MACHINES
Easerva Conversational Booking
A stateful AI-to-human booking application. The server streams model responses over WebSockets while XState workflows, sticky sessions, tools, and persistent storage carry the conversation through discovery and scheduling.
GEOSPATIAL · RENDERING · PRODUCT ENGINEERING
Lyfe Captures Map Studio
Frames an interactive map and sends the exact viewport to the original Python tile engine, producing a native 300-DPI PNG through a one-off Fargate task and signed S3 download.
AI ORCHESTRATION · GITHUB · EVIDENCE ANALYSIS
Candidate Analyzer
Compares a résumé and role against evidence collected from public GitHub projects. A serverless workflow produces a structured report without retaining candidate data.
EARLIER WORK
Selected machine-learning projects.
Seizure Detection
A deep recurrent neural network using extracted body landmarks to identify seizures in video, with transfer learning and a purpose-built training dataset.
View repository ↗Skin Cancer Detection
A Raspberry Pi prototype and convolutional neural network classifying dermoscopic images across seven lesion types using the HAM dataset.
View repository ↗Financial Sentiment Analysis
A domain-specific classifier for financial documents using association rules, performance indicators, and a purpose-built financial lexicon.
View repository ↗CertificationAWS Solutions Architect – Associate
CertificationAWS AI Practitioner – Foundational
EducationTexas A&M · B.S. Computer Science