I automate the cloud and build AI agents that give businesses their time back.
AWS-certified, based in Chicago. I write Python and Terraform, ship through automated pipelines, and design AI agents with guardrails. AI-first is how I work: I build, test and own everything that ships.
LET'S CONNECT
My path is economics → code → systems → cloud & AI, and each step added a skill I still use. Today I automate the cloud with Python, Terraform and GitHub Actions, and build AI agents with guardrails. I'm completing an M.S. in Information Technology at Illinois Tech, and I treat every project as a chance to build something technical and a little bit beautiful.
Fix: Terraform. The bucket, CDN, DNS record and certificate are code, with state stored remotely in S3.
Fix: GitHub Actions runs plan on every pull request and apply on merge.
Fix: OIDC with two least-privilege IAM roles. Nothing to leak and nothing to rotate.
Fix: an AI agent reads every Terraform plan, flags destroys, IAM and open ports, and explains the change in plain words.
Fix: hands-on Linux, DNS, firewalls and Active Directory experience, plus a master's degree focused on cloud and operating systems.
Fix: AI as a daily pair. It drafts scripts, explains errors and reviews changes, and I verify everything before it ships.
AI-first, to me, is a discipline of delegation: machines own execution, while I retain judgment, accountability and taste. Every agent I build operates within explicit guardrails, and every output is verified before it reaches production.
Claude helps me draft Terraform and scripts, explain errors and think through designs. I read, test and own every line that ships.
Every infrastructure pull request on this site is reviewed by an AI agent that explains the Terraform plan in plain language.
The agent can only comment. It has a read-only role, deterministic risk checks, masked account IDs, and it treats the plan as untrusted input.
I'm growing from cloud and DevOps toward automating operations with AI, and toward the infrastructure that machine learning runs on.
Every part of this — the bucket, the distribution, the DNS record, the certificate — is defined in Terraform with remote state in S3. GitHub Actions runs plan on every pull request and apply on merge, authenticating to AWS via OIDC rather than long-lived credentials.
On each pull request that touches terraform/, a read-only role runs terraform plan and the workflow posts one comment: a change table, deterministic risk flags (destroys, IAM, anything open to 0.0.0.0/0) and an AI summary. The agent can only comment, never apply.
Hiring for cloud, automation or AI-first work in Chicago? Say hi, or read the code behind this page.