Scott Campbell

Scott Campbell

Data Science · Statistical Analysis · Applied AI · Decision Dashboards

Download PDF Download Word

Summary

Data scientist with 10+ years of operational ownership and hands-on work in SQL, statistical analysis, optimization, recommendation systems, and evaluated AI. I build and deploy analytical applications, test them against credible baselines, and publish both the results and their limits. My current work includes SQL models and decision reporting for a high-volume manufacturing line. Earlier, I founded and operated a behavioral health practice that grew to approximately 25 employees. M.S. Data Science at Eastern University, in progress.

Technical Skills

Data Science & Statistics

Python, pandas, NumPy, SciPy, statsmodels, scikit-learn, exploratory analysis, statistical inference, forecasting, experimental design

Applied AI & Machine Learning

Recommendation systems, embeddings, hybrid retrieval, text-to-SQL, model evaluation, AI guardrails, citation design, FAISS, sentence-transformers

Data, SQL & Dashboards

SQL, PostgreSQL, DuckDB, dimensional modeling, data validation, dashboard design, Power BI, DAX, Power Query

Delivery & Supporting Tools

FastAPI, REST APIs, Git, GitHub Actions, Linux, nginx, HTML, CSS, JavaScript

Experience

Manufacturing Engineering Technician II

Dec 2023 - Present

First Solar, Perrysburg, Ohio

  • Build SQL data models, analyze maintenance and production data, and deliver Power BI decision reporting for a high-volume manufacturing line.
  • Translate maintenance and production workflows into structured data capture and automated reporting, and present technical findings and recommendations to senior leadership.
  • Maintain and troubleshoot 37 industrial robots and over 100 conveyor systems on the same line, working directly inside the process the analytics describe.

Owner / Operations Lead

Sep 2018 - Nov 2023

Private Practice (Self-employed), Behavioral Health Practice, Ohio

  • Founded and operated a behavioral health practice that grew to roughly 25 employees, owning operations, staffing, scheduling, billing workflows, compliance processes, and financial reporting.
  • Built spreadsheets, tracking systems, and internal reports used to run day-to-day operations and administrative workflows.
  • Managed vendor systems, documentation workflows, payroll-related data, and billing coordination across the practice.

Inpatient Therapist

2016 - 2018

ProMedica Toledo Hospital Psychiatric Unit, Ohio

  • Managed high-volume clinical documentation, treatment planning, and reporting within hospital electronic systems.
  • Coordinated care information across a multidisciplinary team while maintaining accurate, compliance-sensitive records and deadlines.

Social Worker / Data Systems Resource

2013 - 2016

CCNO (contracted), Social Services Agency, Ohio

  • Served as the department's informal data and systems resource while providing social work and case management services.
  • Designed and built Microsoft Access database applications to track cases, manage program data, and automate reporting, replacing manual spreadsheets and paper processes.

Selected Independent Projects

Dispatch Optimizer

Optimization · Live demo · Source

Constraint-based field-service planner using OR-Tools CP-SAT, FastAPI, PostgreSQL, and Next.js. Benchmarked across 200 randomized synthetic scenarios against a greedy nearest-qualified dispatcher with a 2-opt improvement pass, reporting medians with 5th-to-95th percentile bands. The comparison showed the shipped eight-second solver budget sits below the threshold where constraint programming beats the heuristic, which is the result I published.

Black Box AI

Applied AI · Live demo · Source

Natural-language analysis over 7,462 NTSB reports, routing quantitative questions to guarded SQL and narrative questions to cited retrieval. Answers expose queries or source records, while a labeled exploratory evaluation documents retrieval behavior and limits.

arXiv Recommender

Machine learning · Live demo · Source

Five recommendation approaches benchmarked on a 28,000-paper OpenAlex snapshot, with held-out ranking metrics, bootstrap confidence intervals, FAISS serving, and top-10 responses under 100 ms p95.

Fair Lending Lab

Statistical inference · Live demo · Source

Statistical screening platform over 210,643 raw HMDA records, with pre-specified hypotheses, effect sizes and confidence intervals, multi-method sensitivity checks, multiplicity correction, and explicit causal limitations.

Education

M.S., Data Science

Eastern University, in progress

M.S., Information Systems Management (Business Intelligence Specialization)

University of Arizona Global Campus, 2025  ·  GPA 4.0

M.S., Social Administration (MSSA)

Case Western Reserve University, 2016  ·  GPA 3.7  ·  Certificate in Nonprofit Leadership & Management

B.A., Social Work (Minor in Psychology)

Lourdes University, 2013  ·  GPA 3.8  ·  Phi Alpha, Kappa Gamma Pi, Psi Chi

Three completed degrees and an M.S. Data Science in progress span analytics, information systems, organizational leadership, and social work. The mix is the point: data systems are organizational systems, and the technical layer only creates value when it fits the decisions and constraints around it.