I design and build high-leverage software systems: cloud architectures on AWS, automation that removes toil, and AI workflows that put LLMs to work inside real engineering processes. My background is enterprise cloud architecture, so the things I build are secure, cost-aware, observable, and tested, whether that's a serverless platform, a vulnerability-remediation pipeline, or tooling around a COBOL batch shop.

• 🤖 AI-powered automation - LLM agents that turn tickets into pull requests, and tooling that makes AI coding agents (Claude Code, Copilot, Gemini CLI, Codex) better at using real tools
• 🏗️ Cloud architecture - Multi-region, highly available, fault-tolerant, event-driven designs on AWS
• ⚡ Infrastructure as Code - Terraform, CloudFormation, AWS CDK
• 🔄 Serverless & edge - Lambda, API Gateway, DynamoDB, EventBridge, Step Functions, Cloudflare Workers
• 🧱 Legacy modernization - Modern tooling, observability, and dev workflows around mainframe COBOL
• 🛠️ DevOps & CI/CD - GitHub Actions, Terraform Cloud, Harness, Jenkins
• 🔒 Security & compliance - Zero Trust, automated security controls, SOC2/ISO27001
• 💰 Cost optimization - FinOps practices, resource optimization, cost governance
• 📊 Observability - CloudWatch, X-Ray, Prometheus, Grafana, ELK

Modern engineering around mainframe COBOL

A realistic COBOL batch shop (JCL job streams, copybooks, EBCDIC packed-decimal files) that runs in a container without a mainframe, surrounded by the tooling production shops rarely have: job visibility, data interfacing, monitoring, and a modern development loop.

Tech: Rust · GnuCOBOL · JCL · Docker · GitHub Actions

From vulnerability finding to pull request, automatically

Two pipelines: one turns ServiceNow application-vulnerability findings into Jira issues; the other has an LLM read the target repo, write the fix, and open a GitHub PR, then notifies the team over Teams, Slack, and email. Pluggable LLM providers (Claude, Gemini).

Tech: Python · Anthropic Claude API · Google Gemini · Jira · ServiceNow · GitHub API

Teach AI coding agents to use your tools

One catalog of "how to use this tool" knowledge for CLIs (gh, kubectl, aws, ...) and HTTP-only systems (Jira, Confluence, Slack, Splunk, Jenkins, Harness), compiled into the context format each agent reads: Claude Code Skills, AGENTS.md, GEMINI.md, and copilot-instructions.md.

Tech: Python · Claude Code · GitHub Copilot · Gemini CLI · Codex CLI

A full event-driven AWS app, deployed and tested on a laptop

Proof of concept evaluating fakecloud as a local AWS stand-in: ~70 Terraform resources (Lambda, API Gateway, SNS/SQS with DLQs, DynamoDB Streams, S3, Kafka, CloudFront, Route 53) plus a Next.js front end, verified by 22 end-to-end tests, with 15 documented emulator findings.

Tech: Terraform · AWS Lambda (Node.js) · SNS/SQS · DynamoDB · Kafka · Next.js

Used-vehicle opportunity screening

Pulls real listings every 6 hours, normalizes them by VIN, and gates every vehicle on its history (accident, salvage, flood) before ranking it on price, mileage, and NHTSA recall data. Runs entirely on free tiers.

Tech: TypeScript · React · Cloudflare Workers · Hono · D1 · GitHub Actions

Listen to the world's radio on a live map

5,000+ geo-located stations on an interactive clustered map, with HLS playback, a regional "radio scanner" mode, and shareable station lists.

Tech: TypeScript · Vite · Leaflet · HLS.js

Specializing in designing and implementing:

• 🏗️ Multi-Region Architectures - High availability and disaster recovery
• ⚡ Serverless Solutions - Event-driven, auto-scaling applications
• 🔒 Security-First Design - Zero Trust, defense in depth
• 💰 Cost-Optimized Solutions - FinOps and resource optimization
• 📊 Observability - Comprehensive monitoring and alerting

More on the blog at roberjo.github.io.

• Serverless Architecture Patterns for Enterprise Applications - Scalable patterns for modern applications
• Cost Optimization Strategies for AWS Infrastructure - Practical FinOps implementation
• Implementing Zero Trust Security in AWS - Security-first architecture design
• Production Ready Microfrontend Build with Codex - Building a React portal on microfrontends with OpenAI Codex

I'm passionate about helping organizations get real leverage from cloud and AI. Whether you're looking to:

• 🤖 Put LLMs and coding agents to work inside your engineering workflow
• 🏗️ Design scalable cloud architectures
• 🔄 Implement DevOps practices
• 🔒 Enhance security posture
• 💰 Optimize cloud costs

Let's talk.