GenAI & Machine Learning
Engineering Leader
Twenty years building production AI at Fortune 50 scale. Architect of Verizon's first production ML model and first enterprise GenAI platform — with outcomes that reach the Board of Directors.
I built Verizon's first production machine learning model in 2017. Today, I architect the GenAI platforms that put AI in the hands of 900+ HR leaders — and shape strategy at the Board of Directors level.
As Principal of AI & Data Strategy at Verizon, I lead enterprise GenAI initiatives that democratize analytics for hundreds of stakeholders. My work spans the full AI stack — from production LLM platforms on Vertex AI and Gemini, to 95%+ accurate predictive models, to NLP pipelines analyzing 63,000+ employee responses at scale.
What I do best is translate technical capability into outcomes the C-suite can measure. Three consecutive Exceeding/Top Performing ratings. A Brandon Hall Gold Award for AI Innovation. And the trust of senior leadership to present AI strategy at the Board level.
Architected Verizon's first enterprise GenAI platform using Google Vertex AI + Gemini
Top Performing rating in 2025 — third consecutive year rated Exceeding/Leading
Accelerated model deployment from 8 weeks to 3 — while mentoring the next generation of data scientists
Production systems shipped. Real metrics. Quantified business value at Fortune 50 scale.
Enterprise GenAI HR platform delivered real-time, natural-language insights to 900+ stakeholders.
35-point improvement through novel LLM debiasing pipeline, with insights presented to Board of Directors.
Voluntary turnover prediction model (PyTorch + Prophet) enabling proactive workforce planning.
RAG-powered AI summarization agent enabling instant policy retrieval across Verizon's entire HR document library.
Pioneered cause-and-effect analyses that tripled e-commerce close rates for Fios broadband.
Built an AI coding assistant agent (LangChain + Gemini) that accelerated team development speed.
Novel network modeling solution for Smart Scheduler optimized in-office collaboration.
AI Risk Clustering Engine (XGBoost + SHAP) auto-flags Code of Conduct violations, triggering personalized interventions for Legal.
"Scott's GenAI leadership has transformed how Verizon leverages AI for people decisions."
Carlos Moreno Director, HR Operations, Verizon
Full-stack AI leadership — from research design and production deployment to enterprise governance.
Enterprise GenAI platform development using Gemini, OpenAI GPT models, and Llama. RAG pipelines, agentic workflows, prompt engineering, and fine-tuning at scale.
Predictive modeling, classification, clustering, and time series forecasting. Building production ML systems with 95%+ accuracy for workforce analytics.
Large-scale text analysis across 63,000+ employee survey responses. Topic modeling, sentiment analysis, and debiasing pipelines that improved classification accuracy by 35 percentage points.
Rigorous research design, causal inference, A/B testing, and ROI measurement. Quantifying the impact of HR programs with academic-grade methodology.
Building interactive dashboards and self-service analytics platforms. Migrating legacy systems and enabling 1,000+ users with natural language data exploration.
End-to-end model lifecycle management with GitLab CI/CD, Domino, and Airflow. Bias detection, drift monitoring, and ethical AI frameworks.
Multi-cloud platform deployment across GCP, AWS, and Azure. Container orchestration, scalable data pipelines, and enterprise-grade infrastructure for production AI workloads.
Open-source tools and experiments spanning MLOps, agentic AI, edge computing, and security.
Production-grade MLOps pipeline with automated drift detection, self-healing retraining, and feature store integration.
Autonomous research agent that plans, searches, synthesizes, and produces cited reports using LangGraph.
Real-time computer vision assistant optimized for edge deployment with ONNX runtime and voice interaction.
Knowledge-graph-powered RAG system for navigating complex regulatory and compliance documents.
AI-driven security operations center agent with automated threat triage, enrichment, and human-in-the-loop escalation.
Interactive Streamlit app for exploratory data analysis, statistical testing, and visualization — fully offline-capable.
Personal digital assistant integrating Twilio, Claude, and Telegram for SMS-driven AI conversations.
Automated vulnerability assessment tool with CIS benchmark checks, pen testing modules, and anomaly detection.
Real-time speech and text emotion analysis using TensorFlow.js with a React-based interactive dashboard.
ML-gated crypto trading bot using XGBoost for signal confidence scoring and Coinbase API integration.
Offline-first AI coding assistant powered by Ollama and local LLMs for private, air-gapped development.
Cross-platform CLI framework to design, build, and deploy AI agents with 6 LLM providers, 5 agent patterns, DAG workflows, and MCP compatibility.
Southern Methodist University (SMU)
2017 | Dallas, Texas
Dallas Baptist University
2014 | Dallas, Texas
Professional Certificate | Coursera
2020
Cohort Lead & Participant | Verizon & Coursera
3 Certifications, 2020
LinkedIn Learning
If you're shipping ambitious AI at enterprise scale, let's talk.
scott@scottseverance.net
/in/scott-severance
Dallas-Fort Worth, Texas