Joe Caffrey

Enterprise AI architect and transformation leader with healthcare technology experience since 2017, spanning AI governance, knowledge systems, support operations, product, and workforce enablement. Chairs enterprise AI governance and translates ambiguous workflows into bounded, source-grounded AI capabilities with explicit permissions, evaluation, human decision rights, operational ownership, and value measures. Combines enterprise process architecture with hands-on fluency in retrieval, integrations, access controls, agent workflows, and operational resilience.

Digital Content Product Manager - TeleTracking - Pittsburgh, PA

(2024-06)

Concurrent enterprise leadership responsibility: Chair, AI Steering Committee

• Chairs governance review for a portfolio of 200+ AI use-case entries and 20+ connector, GPT, and agent proposals across 9 functions; links intake to value and risk review, security and privacy, human accountability, monitoring, escalation, and business ownership.
• Delivered a responsible-use enablement path reaching 100 unique employees across 20 live sessions; made training a prerequisite for approved AI access and designed a 5-tier competency model aligned to escalating capabilities and controls.
• Designed capability-, application-, workflow-, and governance-based permission models that separate durable SCIM eligibility from product roles, data access, runtime identities, and higher-risk administrative authority.
• Mapped 500+ task and workflow candidates across 10 business domains and created 20+ current- and target-state process maps, including Prospect to Cash, Development Discovery to Launch, and Feature Launch to Enablement.
• Defined target-state composable AI patterns spanning headless interaction, governed MCP/API coordination, RBAC-secured semantic and data layers, permission-filtered retrieval, task routing, human approval, audit evidence, rollback, and accountable ownership.
• Framed a data-democratization strategy to consolidate scattered systems into warehouse-backed semantic and API layers with role-based access, source traceability, and clearer operating truth.
• Built outcome-first portfolio and value models that rank opportunities by operating impact, readiness, risk, adoption effort, and implementation cost while separating modeled capacity and cost avoidance from realized savings.

Conversations AI Knowledge Manager / Client Support Knowledge Manager - TeleTracking - Pittsburgh, PA

(2021-01 - 2024-06)

• Consolidated and rationalized product knowledge spanning approximately 30 years and 30+ on-premise and cloud versions; reduced 5,000 articles to roughly 1,000 core/legacy and 200 emerging-solution sources.
• Operated Intercom/TeleBot knowledge and AI-assisted support workflows serving 3,000+ annual interactions with approximately 60% AI-agent resolution for eligible interactions before case creation; contributed, alongside broader team and process changes, to reducing active case inventory from approximately 300 to 50 over three years.
• Led knowledge lifecycle and platform operations across repository migration, ingestion gaps, content standards, SME validation, quality monitoring, prompt refinement, privacy and identity concerns, data export, duplication, translation, and ticketing integration.
• Built onboarding and staff-development systems using structured courses, interactive exercises, product tours, quizzes, shadowing, coaching, and collaborative case review.
• Developed PowerShell and MongoDB scripts plus data mappings to integrate company and contact records into chatbot and data-management workflows across disconnected systems.

Technical Support Engineer II - TeleTracking - Pittsburgh, PA

(2017-01 - 2021-01)

• Earned two promotions and three special-merit bonuses in approximately three years through high-impact troubleshooting, process improvement, and client support.
• Selected for three onsite high-impact troubleshooting engagements; led severe-issue calls, represented TeleTracking during escalations, and assumed Technical Account Manager responsibilities.
• Managed certificate-renewal operations for two years and created PowerShell-based file verification to reduce manual hotfix validation; also improved ownership workflows, server records, procedures, and mobile troubleshooting.

Graduate coursework - Managing AI Systems - Carnegie Mellon University / CMU Online

B.M., Music Theory & Composition; B.A., Music Performance (Percussion) - Music - The College of Wooster (2016-01)