GOVCIO • AI DATA ANALYSIS & PROJECT MANAGEMENT
AI Data Analysis & Project Management
Validate evidence, define useful measures, connect program state to leadership decisions, and keep humans in control of AI-assisted work.
High-consequence validation, measurement, program leadership, and source-governed AI workflows applied to decision support.
85 / 57
validation tasks / deficiencies
3 / 17
management-control tools / functional checklists
23 / 7
exercise objectives / discrepancies
CASE 01
85-task validation / 57 deficiencies
Worked with the U.S. Department of Energy to review two nuclear technical orders, executed 85 proficiency tasks, identified 57 deficiencies, and authenticated Broken Arrow procedures.
An Air Force Commendation Medal citation credits the work with leading to the first Broken Arrow response-procedure update in 20 years.
CASE 02
AI evidence governance and failure recovery
Built and use multi-model, human-directed AI workflows with standardized briefs, externalized project state, decision/activity logs, source-of-truth controls, structured handoffs, conflict/uncertainty flags, verification gates, and explicit human approval for consequential decisions.
A wrong-source regression was converted into standing source-identity, completion-state, correction-record, and stop-and-verify controls.
CASE 03
Measurement design that exposes blind spots
Managed formal QA and self-assessment across three management-control tools and 17 functional-area checklists. As Wing Inspection Team lead for six exercises, developed 23 Desired Learning Objectives and identified seven discrepancies.
For Moonroot, designed a service-area research method that supplements topline medians with ALICE, child-poverty, school-lunch, SNAP, food-insecurity, labor-force/disability, and tract-level indicators when averages can hide need.
CASE 04
Executive decision support
Coordinated four live detonations and modeled airfield damage from regional threats to give wing leaders a risk and recovery picture.
In Moonroot, built spreadsheet-based decision support combining comparable evidence, valuation segmentation, financial scenarios, operating constraints, risk thresholds, walk-away criteria, and sequenced next actions.
Supporting evidence
- Superintendent scope: 18 Airmen, 19 deployment equipment packages, $6.9M contingency equipment, $300K O&M budget, and 128 purchase orders.
- 64 intelligence reports → 57 realistic training devices → 24 operations / 34 personnel.
- Managed two duty sections, three administrative accounts, 132 technical orders, and 25,000 classified documents.
- Managed the AFRL-linked ATAK server build and coordinated users/requirements across multiple emergency-response agencies; platform interoperability improved across three wing agencies.
- Seven-agency interoperability program with five disaster-response exercises.
Scope & boundaries
My analytical work centers on validation, measurement design, program visibility, and decision support rather than business-intelligence development. Reporting has been built in Excel, structured documents, and briefings.
Operating pattern
Define the decision → Standardize evidence → Check for blind spots → Validate data → Track program state → Surface risk → Translate insight → Recommend action → Measure outcome → Improve the system.
Continue the conversation: Email Devin • LinkedIn
