Implemented academic lab
AWS Transaction Integrity Lab
A controlled path for transaction writes, backed by IAM policies, input validation, and four documented test cases.
Explore the portfolio
01 — 0301 / PROJECTS
Projects & labs.
A closer look at what I build, test, and learn.
Selected work
More projects coming soonWorking project · validation planned
Hybrid AI Lab for Security and IT
A working local-and-hosted AI workflow, with a documented design for security research, IT tasks, and controlled lab use.
PROJECT 01 / CLOUD SECURITY
AWS Transaction Integrity Lab.
Applying Clark-Wilson concepts with Lambda, S3, and IAM
The problem to solve
Protect transaction records from direct modification while allowing validated writes through a designated procedure. The lab translates a formal integrity model into a small, testable cloud architecture.
- Transaction input
- Lambda validation
- S3 record storage
The execution role authorizes the intended write path. A bucket policy denies direct writes by other principals. CloudWatch and CloudTrail support logging and audit review.
Controls implemented
Validate before writing
A Python Lambda function checks required fields, positive numeric amounts, and distinct submitter/approver values before writing a transaction record.
Limit the write path
The execution role grants S3 write access to the designated objects and CloudWatch logging, without S3 read or delete permissions. An explicit bucket-policy deny restricts direct writes by other principals.
Preserve history and audit evidence
Configured S3 public-access blocking and versioning, plus a multi-region CloudTrail trail with management events and relevant S3 data events. Enabled log-file validation to generate digest files for later integrity checks.
Four documented test cases
The lab report records the following outcomes. The tested write path remained usable while direct uploads were denied.
01
Valid transaction; distinct submitter and approver values
Accepted
Accepted; record written to S3
02
Matching submitter and approver values
Rejected
Rejected; function returned a 403 response value
03
Direct S3 upload using an administrator identity
Blocked
AccessDenied
04
Valid Lambda write after applying the bucket deny
Accepted
Accepted; intended write path remained functional
Scope of the project
The approval rule compares supplied values; it does not authenticate a second approver. This is a Clark-Wilson-inspired lab, not a complete implementation or a compliance assessment. Versioning and CloudTrail support recovery and audit review, but do not by themselves provide immutable retention or verify the validity of stored records.
What I learned
The key lesson was how application validation, identity permissions, and resource policies reinforce one another. Testing the denied path was as useful as testing successful writes: it showed that the applied bucket policy blocked a direct upload while preserving the intended Lambda workflow.
Next steps
- Bind approval to independently authenticated identities.
- Add a procedure that checks stored-record integrity.
- Evaluate protected record/audit retention and alerts for denied writes.
PROJECT 02 / AI & SYSTEMS
Hybrid AI Lab for Security and IT.
Local and hosted models for technical research and development
The problem to solve
Create a flexible environment for technical research, coding assistance, and troubleshooting. Local and hosted models provide different options, with the choice guided by the task, available resources, and the information involved.
- Local + hosted models
- Human-reviewed workflow
- Scoped lab environment
The guide connects model services with chat or development tools and a separate Linux lab environment. The full architecture and its controls are design work; end-to-end validation remains a next step.
Current work
Local model deployment
Set up local language models for hands-on use in technical research and IT workflows.
Hosted model integration
Combined local models with hosted AI services to support technical tasks and AI-assisted application and website development.
Lab documentation
The project guide covers model services, user interfaces, development workflows, and a proposed virtualized security lab, recording the intended operating approach.
Design choices and scope
The local-and-hosted workflow is in use. The broader security-lab architecture is documented; benchmark results and security-control validation are not yet included.
Deliberate model selection
Choose local or hosted processing based on the task and review what information would leave the local environment.
Human review of actions
Keep model suggestions separate from execution. Review generated code and proposed commands before using them in a scoped lab.
Separate lab resources
Design for dedicated lab targets, separate credentials, and recoverable virtual environments. Keep secrets out of shared documentation.
Scope of the project
The current scope is model deployment, integration, AI-assisted development, and documented lab design. Validation of environment isolation, model performance, and adversarial behavior remains future work.
What I learned
The value of the project is in connecting tools into a usable workflow and making the design understandable. Model output still needs review, testing, and clear documentation before it becomes a dependable technical result.
Next steps
- Verify network boundaries and test virtual-machine snapshot recovery.
- Validate credential separation and the information sent to hosted services.
- Run a repeatable set of benign coding and IT tasks; document accuracy, latency, and required human corrections.
02 / RESUME
The short version.
Education, technical skills, and experience in one place.
Technical skills.
Practical experience.
I combine hands-on work in Windows, Linux, networking, and AI-assisted development with more than five years of management experience.
EDUCATION
B.S. in Cybersecurity
Bellevue University
Expected December 2026 · GPA 3.84
Select the document to open the full resume
SYSTEMS & NETWORKS
Active Directory, Group Policy, Windows, Linux, TCP/IP, and DNS fundamentals.
AI & DEVELOPMENT
AI-assisted applications and websites, local LLM deployment, and hosted model integration.
LEADERSHIP
Team coaching, employee training, customer problem resolution, and daily operations.
03 / WORK EXPERIENCE
Experience in practice.
The roles, responsibilities, and lessons that shape my work.
Apprentice
Assistant General Manager
- Lead daily restaurant operations, coach team members, and coordinate priorities during high-volume service.
- Train employees on operational procedures and resolve customer concerns with clear communication and follow-through.
Operations · Team leadership · Employee development
Service Manager / Service Leader
- Coordinated shift assignments and coached employees on service, operational procedures, and safety standards.
- Addressed customer and operational issues while supporting team members during busy service periods.
Shift coordination · Coaching · Customer service
Crew Member
- Supported customer service and daily restaurant operations.
- Promoted to Service Manager in May 2021.
Teamwork · Service · Daily operations