Peter Shull
Principal AI Systems Consultant | Agentic AI Systems | Domo AI & Enterprise Analytics
15 yrs experience · North Augusta, SC · <10 hrs/week
About
Principal AI Systems Consultant and PharmD with 15+ years bridging clinical operations, enterprise analytics, and AI systems engineering. Designs governed agentic AI systems for inventory QA, operational analysis, and executive decision support—hands-on across Domo AI, React/TypeScript, JavaScript, Python, SQL, DuckDB-WASM, semantic contracts, text-to-SQL validation, MCP-style tool workflows, and eval harnesses, translating messy operational reality into auditable AI systems executives and operators can trust.
Skills
Experience
- Principal AI Systems Consultant, Agentic Inventory QA and Analytics · Clearview BI / StartDayOne01-01-2026
Principal AI Systems Consultant, Agentic Inventory QA and Analytics • Hired to design and build governed agentic AI systems for inventory QA, operational analysis, and executive-facing analytics in Domo-based enterprise environments. • Built Domo Executive Inventory Agent and Domo AI QA Bot using Domo AI, @domoinc/query, Domo Query APIs, AppDB, DuckDB-WASM, React/TypeScript, JavaScript, semantic contracts, governed fetch plans, and source-grounded answer synthesis. • Re-architected AI execution from direct AI-generated SQL toward a safer fetch-plan pattern where AI proposes scoped data requests, Domo pushes down filtering/aggregation, DuckDB-WASM performs local joins, and validated evidence rows drive final responses. • Implemented semantic-contract enforcement across governed datasets and approved joins, constraining AI behavior to approved aliases, fields, business definitions, join keys, date windows, row budgets, filter scope, and detail-table gates. • Built AST-based SQL safety controls using pgsql-ast-parser and node-sql-parser patterns to reject DDL/DML, unsafe operations, unfetched tables, unapproved joins, current-date drift, and non-SELECT execution paths. • Designed detail-table gating, row-budget controls, deterministic fallback logic, refusal behavior, active-filter propagation, admin override controls, audit logging, schema diagnostics, and production-style health panels. • Modeled NetSuite source extracts into a 21-dataset Domo architecture with dimensions, facts, metric tables, and gated detail layers for dashboard performance and AI-agent query reliability. • Built inventory QA and analysis logic for stockout risk, days of supply, velocity windows, vendor lead time, open demand, trapped capital, price variance, fulfillment SLA, customer activity, and exception prioritization. • Created reusable Domo Custom Apps and MCP-style development frameworks covering React/Vite scaffolding, manifest validation, AppDB, Code Engine, dataset schema verification, Pro Code publishing, page/card automation, testing, logging, and deployment. • Built a Claude/Codex/Gemini engineering harness with routing rules, coding-agent guardrails, smoke tests, failure logs, evidence logs, review templates, plugin audit checklists, and sensitive-data boundaries. • Produced client-facing deployment guides, semantic contract documentation, metric formula references, smoke-test checklists, governance instructions, troubleshooting notes, and executive-ready implementation plans.
- Principal Consultant and Data Architect · Clearview BI / Argo Analytics05-01-2025 – 12-01-2025
Principal Consultant and Data Architect • Consolidated 30 disparate scheduled processes into a single optimized workflow, reducing compute cost by 96.7% and eliminating data freshness delays. • Built reusable BigQuery and Domo analytics components for healthcare finance clients, partnering with customer success, engineering, and client stakeholders to define scope and delivery standards. • Wrote Python and SQL automation for data ingestion, validation, data contracts, SLA monitoring, and error detection before client impact. • Reverse-engineered legacy BI logic and rebuilt calculations, transformations, and dashboard experiences to preserve KPI parity during Domo migrations.
- AI Engineer and Researcher · Independent AI Research and Development08-01-2024 – 12-01-2025
AI Engineer and Researcher • Built privacy-first local AI prototypes, including a low-latency voice pipeline using Whisper, Qwen 2.5, and RVC on consumer GPU hardware without sending sensitive data to cloud services. • Developed citation-constrained RAG prototypes over FDA DailyMed data for clinical drug-information questions, applying clinical safety standards to response validation and hallucination control. • Designed evaluation frameworks for prompt reliability, source grounding, refusal behavior, and clinical-output safety before applying the same patterns to enterprise analytics agents.
- Business Intelligence Program Manager · BlueBin Inc.03-01-2023 – 07-01-2024
Business Intelligence Program Manager • Led migration of 70+ dashboards from Tableau to Domo, replacing fragmented spreadsheets and siloed reports with governed data models. • Implemented analytics model versioning and documentation practices in GitHub to improve auditability, rollback readiness, and team handoff. • Built a Python templating API that reduced dashboard setup from approximately 8 hours to 30 seconds, a 99% efficiency gain, while eliminating manual copy-paste errors. • Built a Microsoft Power App for inventory audits that increased data submission rates by 40% and closed a downstream analytics data-quality gap. • Trained clients through analytics best practices and a self-service training portal, increasing user adoption by 80%.
- Senior Data and Project Manager · Premier Health Network06-01-2021 – 09-01-2022
Senior Data and Project Manager • Built the centralized analytics function from scratch for a clinically integrated network managing 65,000+ Medicare Advantage lives across primary care and behavioral health domains. • Integrated data from 36 disparate EMR systems and a Health Information Exchange into a single core data model, resolving major interoperability blockers across HL7 and CCD data. • Audited CMS reimbursement files using claims and CPT-code expertise, identifying coding errors that recovered $160,000 in lost revenue. • Delivered provider-facing dashboards that drove a 20% improvement in care-gap closures and a 15% increase in reimbursement rates for Medicare Advantage contracts.
- Clinical Pharmacist · U.S. Department of Veterans Affairs10-01-2019 – 07-01-2020
Clinical Pharmacist • Verified high-volume medication data with 99%+ accuracy while maintaining strict adherence to federal safety, privacy, and clinical documentation standards. • Wrote VBA macros to automate processing of 1,500 prescriptions weekly, saving approximately 25 hours per week of manual data entry.
- Pharmacy Manager · Walgreens / BI-LO Pharmacy10-01-2011 – 10-01-2019
Pharmacy Manager • Managed pharmacy operations, inventory controls, clinical workflows, and compliance obligations across HIPAA, CMS, OIG, and state pharmacy-board requirements.
Education
- University of South CarolinaDoctor of Pharmacy (PharmD)
- ThinkfulData Analytics Immersion Bootcamp
- Vanderbilt UniversityAdvanced Prompt Engineering
- PMI / LinkedIn LearningGenerative AI for Product Managers
- LinkedIn LearningAutomate Your Workflows with Generative AI
- LinkedIn LearningTelling Stories with Data
- GoogleGenerative AI for Educators
Similar talent on Pangea
Hire Peter through Pangea
Describe your project to the Pangea agent — see if Peter is a fit, with transparent pricing and interviews booked straight onto your calendar. No contact details change hands until you hire.
See if Peter is a fit