I build applied AI and platform tools that turn ambiguous operational problems into working software: LLM-assisted workflows, agentic tooling, infrastructure automation, cloud integrations, and customer-facing technical prototypes.

My background combines senior enterprise architecture, hands-on software delivery, startup CTO work, and modern AI engineering. I’m strongest where AI systems need to connect to real business context: APIs, cloud platforms, identity, infrastructure, telemetry, runbooks, GitHub workflows, and existing enterprise systems.

• Best fit: Forward Deployed Engineer · Applied AI Engineer · AI Solutions Architect · Senior/Staff Technical Builder
• Core signal: Applied AI systems, agentic workflows, enterprise integration, CLI tooling, Terraform/IaC automation, cloud deployment patterns, customer-facing prototypes
• Recent proof: AI Terraform plan reviewer · LLM-generated cloud estate briefings · Terraform scaffold generator · bounded operations agent · Go-based operator CLIs · deployed cloud reference architectures

GitHub Action for AI-assisted Terraform plan review. Parses Terraform plan output, identifies IAM, network, destructive-change, and sensitive-output risks, and posts LLM-generated PASS/WARN/BLOCK review comments to pull requests.

Why it matters: Shows applied AI in a real engineering workflow: constrained outputs, risk classification, PR automation, multi-provider LLM support, and practical human-in-the-loop review.

Stack: Python · GitHub Actions · Terraform · Anthropic Claude · OpenAI · Azure OpenAI

LLM-powered briefing workflow that converts multi-cloud inventory artifacts into plain-English estate summaries, anomaly callouts, security observations, drift detection, and recommended next actions.

Why it matters: Demonstrates an FDE-style pattern: ingest messy customer or environment data, convert it into an actionable technical narrative, and produce outputs useful to both engineers and stakeholders.

Stack: Python · Anthropic Claude · OpenAI · GitHub Actions · AWS · Azure · GCP

AI-assisted Terraform scaffolding tool that turns plain-language architecture descriptions into infrastructure code with validation, guardrails, and least-privilege defaults.

Why it matters: Shows practical AI-assisted engineering: constrained generation, safety checks, infrastructure patterns, and a path from natural-language intent to deployable technical artifacts.

Stack: Python · OpenAI · Azure OpenAI · Terraform · GitHub Actions

Autonomous operations agent with watchdog, incident response, and briefing modes. Designed around bounded tool use, structured reasoning logs, explicit write gating, mock mode, runbooks, and escalation paths.

Why it matters: Shows agentic system design with operational safety controls: tool boundaries, auditability, human escalation, and repeatable investigation workflows.

Stack: Python · Anthropic Claude · AWS IoT Core · DynamoDB · CloudWatch · structured JSONL logging

Go-based operator CLI for managing fleet status, telemetry, and event workflows through real AWS API calls.

Why it matters: Shows hands-on CLI development, cloud API integration, and operator-focused tooling — useful patterns for forward-deployed and applied AI deployment environments.

Stack: Go · Cobra · AWS SDK v2 · CLI tooling

End-to-end AWS reference implementation showing X.509 identity, MQTT/TLS ingestion, Lambda processing, DynamoDB storage, CloudWatch monitoring, and alerting — deployed and validated against real AWS infrastructure.

Why it matters: Demonstrates the ability to turn a complex operational domain into a secure, documented, repeatable cloud architecture. Useful as proof of integration depth, deployment judgment, and production-oriented design.

Stack: AWS IoT Core · Lambda · DynamoDB · CloudWatch · IAM · MQTT/TLS · X.509

MCP server for AI-assisted enterprise operations. Intended to expose cloud inventory, Terraform plan artifacts, runbooks, architecture notes, and incident context as bounded MCP resources, tools, and prompts.

Planned design:

• Read-only by default
• Explicit write gating
• Tool allowlists
• Structured audit logs
• Mock fixtures for safe demos
• Customer discovery and deployment-risk prompts
• Human escalation paths

Why it matters: MCP is becoming a practical bridge between AI assistants and real enterprise systems. This project will demonstrate how to expose operational context safely and usefully to AI clients.

• Applied AI systems — LLM workflows that connect to real tools, APIs, infrastructure, and operational context
• Forward-deployed engineering patterns — discovery, prototype, integration, validation, handoff, and repeatable deployment assets
• Agentic workflows — bounded tool use, structured logs, explicit write gating, escalation controls, and human-in-the-loop design
• Enterprise integration — APIs, IAM, OAuth/OIDC, SSO, RBAC, Active Directory, cloud platforms, and existing enterprise systems
• AI-assisted DevOps — Terraform review, estate briefings, infrastructure scaffolding, GitHub Actions, and cloud automation
• CLI and operator tooling — Go, Python, Bash, cloud SDKs, Docker, and practical workflows for technical users
• Cloud deployment patterns — AWS, Azure, GCP, Docker-based application delivery, monitoring, and runbooks
• Customer-facing technical demos — translating complex systems into understandable prototypes, briefings, and architecture narratives

Open to senior/staff Forward Deployed Engineer, Applied AI Engineer, AI Solutions Architect, and customer-facing technical builder roles — especially where AI systems need to integrate with real enterprise environments.

LinkedIn