Technical & Agentic Track
Self-paced track · 7 topics · moderate → advanced
Go beyond chat: package reusable instructions, run multi-step and looped tasks, and build assistants that connect to your files and tools. Start from Workshop 5 (agentic basics) and keep Workshop 1’s data rules in view. The throughline: more autonomy for the tool means more deliberate oversight from you.
Difficulty levels
Novice getting started — no prior experience needed. Moderate comfortable with some AI use. Advanced experienced users going deep.
Topics
| # | Topic | Level | You’ll be able to… |
|---|---|---|---|
| 1 | Creating & Using Skills | Moderate | Package a repeated task once and reuse or share it |
| 2 | Context Engineering | Moderate | Give a model the right context so it stays on task |
| 3 | Building AI Agents with Natural Language | Moderate | Assemble an assistant from instructions and tools — no coding |
| 4 | Creating & Using Loops | Advanced | Repeat a task across many items, reviewing each pass |
| 5 | Local Models via Ollama | Advanced | Run open models on your own machine for privacy and cost |
| 6 | RAG & Knowledge Bases over Your Documents | Advanced | Ground answers in your own document set, with citations |
| 7 | MCP & Connecting AI to Data & Tools | Advanced | The plumbing that lets agents reach data sources and tools |
Not sure where to start?
Answer a few questions to find your starting point in this track.
WarningData rules apply throughout
Use public or non-sensitive data for every activity, and match any tool to Drexel’s data classification. See Tools & Access and Drexel’s AI Tools page.