AI Reliability Architect | Python Developer | Simulation Builder | Data Processing Specialist
Contact: sheldonksalmon2026@gmail.com
Portfolio: github.com/AionSystem
DOI Index: 17 Citable Works
I build systems. Over two years of independent engineering, I created the AION Constitutional Stack — 17 frameworks, 9 repositories, 2,040+ files, and a working certainty scoring engine. My work is timestamped, sealed via Zenodo (17 DOIs), and open for review.
I am now seeking freelance opportunities in:
- Web scraping and data extraction
- Data processing, cleaning, and normalization
- Simulation development (Python)
- AI training and validation (physics, math, logic)
- Educational content generation
| Category | Skills |
|---|---|
| Python | Simulation development, data extraction, validation scripts, NumPy, Pandas (learning), SciPy (learning) |
| Data Processing | JSON, CSV, Google Sheets, data cleaning, normalization, structured output formatting |
| Web Scraping | BeautifulSoup (learning), Selenium (learning) — will be proficient within days |
| AI & LLMs | Built FSVE scoring engine; experienced with GPT-4, Claude, Gemini, Grok |
| Version Control | Git, GitHub, Zenodo DOI registration |
| Simulation | Physics engines (roller-coaster, g-forces, energy), magnetometer (UXO detection), math worksheet generation |
- Built a full physics engine for energy, g‑forces, loops, and braking calculations
- Demonstrates Python proficiency, mathematical modeling, and validation
- GitHub
- Built for UNDP innovation competition; tri‑axis magnetometer gradiometer simulation
- Data extraction, normalization, and structured output
- GitHub
- Generates custom arithmetic, fractions, linear equations, and PEMDAS worksheets
- Data cleaning and formatting to CSV/JSON
- GitHub
- Live on PyPI (
pip install sovereign-trace); cryptographic sealing with SHA-256 - 11 sealed ledger entries; Zenodo DOI registered
- PyPI | DOI
- Full-stack platform with offline capability, AI photo analysis (OpenRouter), and responder dashboard
- UNDP Accelerator Lab Prize entry
- Live Demo | DOI
| Role | Organization | Dates | Key Achievements |
|---|---|---|---|
| Founder & AI Reliability Architect | AionSystem | 2024–Present | Built 17 frameworks, 9 repositories, 2,040+ files from scratch without institutional backing |
| Certified Nursing Assistant | Samaritan Keep Home | 2024–2025 | High-stakes precision care — zero tolerance for error in patient-facing decisions |
| Correctional Officer | Belknap County DOC | 2019–2023 | Supervised 100+ individuals daily; developed conflict-resolution protocols |
| Infantryman (11B) | NY Army National Guard | 2011–2019 | Eight years of mission-critical logistics and tactical execution under pressure |
Military service, corrections, and healthcare taught me that protocol failure has immediate human consequence. I bring that standard to AI reliability. Every output I produce is tagged with evidence level ([D] Data, [R] Reasoned, [S] Strategic, [?] Unverified). Nothing exits without a tag. Nothing is presented as more certain than it is.
- GitHub: github.com/AionSystem — 9 brain repositories, all public
- Live Site: aionsystem.github.io — framework hub, 9 pages deployed
- DOI Index: 17 citable works
- Sovereign Trace Protocol:
pip install sovereign-trace| PyPI - VERITAS Demo: Live Platform (Responder code:
UNDP2026)
- Some college: Agrobiology (Japan), Physical Therapy, Criminal Justice (USA)
- Military: Basic Combat Training, Advanced Infantry Training (11B)
- CNA Certification (New York State, Active)
No degree. 17 DOIs. One coherent stack.
- Hours: 10–20 hours per week, flexible
- Rates: Negotiable based on project scope ($25–$55/hour range)
- Contact: sheldonksalmon2026@gmail.com
- Response time: Within 24 hours
"The question is not whether your AI is confident. The question is whether its confidence is earned."