# Oluwatofunmi Badejo

Software Engineer | AI & Data Systems | Cloud & Production Reliability

samuelbadejo928@gmail.com | 401-954-2752 | Providence, Rhode Island

## Summary

Computer Science student who builds, tests, deploys, and evaluates software across web applications, AI, data, cloud infrastructure, and security. I work with Python, FastAPI, Next.js, Phoenix/Elixir, PostgreSQL, Docker, GIS workflows, automated testing, and production-readiness validation. My projects focus on reliable behavior, measurable acceptance criteria, and clear release decisions.

## Work Experience

### AI Trainer: Software & Product Evaluation Expert

**Handshake AI Fellowship | Remote**
Feb 2026 - Present

- Evaluated AI-generated software projects and model outputs for usability, logic, technical accuracy, requirement fit, and production risk.
- Reviewed agent-built applications for product judgment, implementation quality, acceptance criteria, safety boundaries, and release readiness.
- Wrote structured technical feedback for AI systems used in software development and product evaluation tasks while following confidentiality guidelines.

### Community Donations Center Ambassador

**Savers | Providence**
Apr 2023 - Aug 2024

- Managed donor intake, item sorting, donor questions, and donation-area organization during busy operating periods.

## Selected Projects

### GeoVisLM

Python/FastAPI GIS and scientific visualization project for terrain analysis, vector overlays, reports, dashboard runs, and GeoMiniLM workflow evaluation.

- Built terrain workflows for DEM inputs, slope, hillshade, risk maps, output browsing, and report generation.
- Added dashboard support for projects, runs, validated uploads, lifecycle history, authentication, bearer access, and optional PostGIS storage.
- Checked Docker Compose paths with dashboard, worker, PostGIS, persistent output storage, health checks, and readiness checks.
- Kept GeoMiniLM out of production after candidate `0bc44b0` scored `0.6475` against a required `0.7600`.

### SysEthic

Local security-operations workspace for telemetry, incidents, audit history, playbooks, controlled validation events, and approved response actions.

- Completed automated security-critical coverage for authentication, permissions, audit logging, approval floors, route inventory, invalid input rejection, exception sanitization, and fail-closed response behavior.
- Added controlled validation events for CPU pressure, memory pressure, suspicious network activity, and sensitive-file access; kept full acceptance blocked for a P1 issue and remaining manual checks.

### FitTrack

Phoenix/LiveView fitness workspace for workout logging, reusable plans, nutrition, exercise media, workout history, and strength tools.

- Added explicit workout lifecycle states: draft, active, completed, and discarded.
- Added immutable origin snapshots so completed workout history stays stable after plans, exercises, or source templates change.
- Verified lifecycle behavior with `mix precommit` passing 275 tests and production media coverage across 488 templates.

### Command Center

Next.js App Router workspace for agency morning briefings, proposal generation, proposal persistence, and deployment readiness checks.

- Added domain-aware proposal drafts for website, finance, school, medical, and custom project types.
- Checked production `/api/health` in database mode with migrations ready and `demoFallbackAllowed: false`.
- Confirmed a controlled production proposal survived redeployment and left rollback as the remaining readiness item.

## Education

### Plymouth State University

B.S. Computer Science, in progress  
Started Aug 2024

### Community College of Rhode Island

A.S. General Studies  
Aug 2019 - May 2022

## Skills

**Languages:** Python, JavaScript, Elixir, HTML, CSS  
**Frameworks:** FastAPI, Next.js, Phoenix/LiveView  
**Data & Infrastructure:** PostgreSQL, PostGIS, Docker, GIS workflows  
**Engineering:** Git, automated testing, API development, CI/CD, production-readiness validation, acceptance testing  
**AI & Evaluation:** AI software evaluation, model-output evaluation, acceptance criteria, safety boundaries