Session

Building an AI Agent with PL/SQL, Oracle Text and APEX

What actually happens inside an AI agent? In this session we build an AI agent entirely in PL/SQL using Oracle Database 26ai and Oracle APEX, with no external framework. The agent's job is to answer plain-language questions using Oracle Text.

The idea is that the LLM never needs to know about Oracle Text CONTAINS syntax. It simply completes a structured request (must-have terms, alternatives, exclusions) which PL/SQL then validates, escapes and uses to build an Oracle Text query. If a search finds no results, the code broadens it automatically using fuzzy matching, stemming, query relaxation, and similar techniques .

We walk through the search tools exposed to the model, the REST call and agent loop that bring everything together, and the guardrails that help keep an agent behaving as we want it to. We also look at swapping a hosted model for a local one.

By the end, participants should have a better understanding of how agents work and have a simple pattern they can use to experiment with their own data.

Niall McPhillips

♠︎ Oracle ACE Director. PL/SQL devt. ORDS. APEX. Data Models. REST + Web Services. Multimedia. Text. Int'l Organisations. Energy.

Genève, Switzerland

Actions

Please note that Sessionize is not responsible for the accuracy or validity of the data provided by speakers. If you suspect this profile to be fake or spam, please let us know.

Jump to top