The latest technology
Eighteen-plus years at Oracle across AI, cloud computing and data systems. I teach from real-world practice, so students learn the tools and habits they will meet in their first jobs.
Instructor · Guest lecturer · Enterprise architect
Pat Shepherd
Complex AI, cloud and data topics made simple, relatable and engaging, with the real-world lessons that rarely make it into a textbook.
Based in the Maryland / Washington, DC area · Available in person and virtually
What I bring
Real-world experience, in plain English.
I am a passionate, engaging teacher and trainer. I offer impactful learning built on decades of real-world experience, and an ability to take complex subjects and make them simple, relatable and engaging.
Expansive training
I develop and deliver technology training in a range of formats, from live demos and discussion to exercises, plus public, online modules that students can run on their own laptops in class.
Enterprise architecture
Years of applied architecture work, designing and presenting solutions for Fortune 500 clients across industries.
Beyond the textbook
What the slides don't always show.
Technology changes quickly. The judgment behind it lasts longer. Along with the concepts, I share what years of real-world work have taught me, so students leave with context they can use on day one of a career.
Real-life situations
Stories from eighteen-plus years working with some of the largest organizations across industries. Students hear how projects actually unfold: the constraints, the trade-offs and the decisions behind the architecture diagram.
Lessons for early professionals
What I want every new graduate to know before their first job: how to pick up a new technology quickly, ask better questions, and become someone a team is glad to have.
Things to watch out for
The pitfalls that catch smart people: surprise cloud bills, data stuck in silos, ideas that never get funded, and solutions that do not fit the problem they were meant to solve.
How decisions really get made
Behind every technology choice are budgets, risk, people and timing. Students learn to see the business context around technical work, and how to explain their ideas to non-technical audiences.
Training experience & approach
Learning that fits the room.
As Content Lead for Oracle LiveLabs, I helped build a free, global, hands-on learning platform with more than 900 interactive workshops, used by developers, DBAs, architects and customers worldwide. I have also developed internal and public training videos across key topics.
I have presented, trained and demoed at large-scale events including Oracle Data Deep Dive, Oracle AI World Tour, the Gartner Data and Analytics Summit and Informatica World, and delivered training at field bootcamps and to sales development teams.
My style is to start with what the audience needs to hear, then choose the approach that fits the topic and the room: a live demo, a discussion, or an exercise where students work with the concept themselves. That blend was honed across enterprise architecture, integration business development, field technology enablement, competitive analysis and database product management. My experience as a 5th-degree Black Belt Taekwondo instructor adds energy and passion to the classroom.
A wide range of teaching topics
Pick from these ready-to-go topics, or I can create ones for your needs.
These are some of the topics most relevant to students today. Each can be tailored to your course, and if your institution wants a topic that isn't listed, I'm glad to create and structure a single session or a series around it. Many come with a complete teaching deck, and modules marked Live demo include a demo that can run in class, and those marked Live exercises are built around individual and group exercises.
AI frontier
- AI in the real worldLive demo
- A story-driven tour of how AI works and what happens when it meets a real enterprise. Following an architect at ShopSphere, a fictional retailer, students cover what AI is, how machines learn (machine learning, deep learning, language and vision), why data quality and governance decide whether projects succeed, and field lessons such as token maxing vs. token economics and matching the model to the job. Includes a live demo of building a working system with AI.
- Build with agentic memoryLive demo
- Good agentic AI is not a goldfish. Following one customer of ShopSphere, a fictional retail company, through four progressive live demos, students learn how memory carries context across sessions, what to store and what to keep out, how to match data types to the information, and how to save and retrieve memories with Python alongside live product and customer data. Includes a class exercise.
- Artificial Intelligence and Data 101Live demo
- AI as a data problem first: quality, governance and architecture, not just modeling.
Cloud & data fundamentals
- Relational database 101Live demo
- Reliable databases protect transactions and keep operations running. This is where the story starts: setting up the database behind the CRM of ShopSphere, a fictional retail company, the foundation for the applications students build in the other courses. Following its team, students learn what a database does and how it fits into the data pipeline, warehouse and lake, then work with tables, keys, constraints and SQL, running statements in a live database to see how each one behaves. They also see how business requirements become a database design, and how Oracle runs it behind the scenes.
- Database modeling
- How real business requirements become a sound data model, the design step behind databases that stay reliable and fast as they grow.
- Cloud computing fundamentals & economicsLive demo
- How cloud platforms launch faster, scale on demand and improve resilience, paired with cost-aware architecture, TCO and FinOps trade-offs that prevent surprise bills. Concepts apply across hyperscalers.
- Integrating data in the enterprise
- Connecting siloed legacy and modern systems to speed decisions and unlock analytics and AI.
Modern data types
- AI vector searchLive demo
- The foundation of semantic search, recommendations and more accurate RAG experiences, with a live demo students can follow.
- Property graphsLive demo
- Graph models behind fraud detection, recommendation engines and knowledge graphs.
- JSONLive demo
- Nearly every API speaks JSON. Native semi-structured data handling in databases, immediately applicable.
Applied & career skills
- Winning buy-in for technologyLive exercises
- Securing funding and support for a technical idea starts with customer-focused value, a career skill rarely taught in computing programs. Built around individual and group exercises, students develop the skills to pitch their ideas with that value front and center.
- LiveLabs learning 101Live demo
- A tour of an enterprise-scale learning platform with 900+ workshops, with students running labs on their laptops in class.
LiveStack industry experiences
Finance
A paired demo and workshop follow a database administrator at a fictional bank as six teams ask different things of the same finance data: JSON for a web and mobile app, semantic search for products, a graph to trace connected accounts in a fraud review, spatial distance for routing service work, a machine learning model for demand, and plain-language questions turned into SQL. Students see how one Oracle AI Database can serve each need without separate copies of the records, and how privileges, restricted AI profiles and execution history keep AI-assisted work visible and controlled. Along the way they practice weighing business value, risk and implementation trade-offs.
State & local government
A paired demo and workshop follow a database administrator for a state and local government as six teams ask different things of the same public-service data: JSON for a resident-service application, semantic search that finds services related to a resident concern even when the wording differs, a graph to coordinate public programs and community partners, spatial distance to measure access to service centers, a machine learning model for service demand, and plain-language questions turned into SQL. Students see how one Oracle AI Database can serve each need without separate copies of the records or a separate security model for every feature, and how privileges, restricted AI profiles and execution history keep AI-assisted work visible and controlled. Along the way they practice connecting technical choices to service outcomes.
Track record
Sessions and events I have led.
Over the last few years I have led about 30 webinars, workshops, keynotes and enablement programs, for audiences from a few dozen people to thousands.
Public and customer webinars
- Design Workflows with GenAI and RAG to Streamline Approval Processes
- How to Extract Value from Structured and Unstructured Data with Oracle AI Data Platform
- Databasics: Oracle AI Vector Search Basics
- AI-Powered Reporting with Oracle Analytics AI Assistant and LLMs (scroll down for video)
- Improve Sales and Revenue Forecasting Using Oracle GenAI Services (scroll down for video)
- How to Build Smarter Search Using GenAI Agents (scroll down for video)
- Eliminate Abandonment with GenAI-Powered Project-Based Shopping Carts (scroll down for video)
- AskFinance: Use GenAI with ERP Data (scroll down for video)
Keynotes and events
- Keynote Converged database as the foundation for enterprise AI, for a financial-services customer tech day (100+ attendees)
- Industry event Gartner Data and AI Summit (about 5,000 attendees)
- Booth Oracle Cloud World Tour, Austin: Oracle Database 23ai (50+ customer booth interactions)
- Booth Informatica World: database and AI demos
Sales and field enablement
- GenAI and multicloud workshops for field teams in Burlington and Santa Clara (130 to 300 attendees)
- Hands-on GenAI and machine learning workshops and training programs, including a grading rubric I built
- A database consolidation on Exadata session at a software bootcamp (about 600 attendees)
- Multicloud pitch, training and certification for sales development reps and their managers (60 to 200 attendees)
- Competitive-landscape sessions, including content review and hosting for about 380 sales colleagues
- Autonomous data management, intelligent document processing, modern data platform for retail, and MySQL HeatWave
Internal sessions; recordings are not public.
Work together
Let's put something on the calendar.
I am happy to tailor a session to the topics and format that work best for your courses: a single guest lecture or a recurring slot across the semester. My home region is the Maryland and Washington, DC area, where I'm happy to join you in person, and I'm glad to teach virtually anywhere.
Send a connection request with a note about your course, and I'll be in touch.