How to Run LLM-Generated SQL Safely in Oracle
Stop a language model's SQL from changing or revealing data with plan checks and a read-only schema, and measure how often it is right.
How to Generate SQL from Natural Language with an LLM in Oracle
Let users ask questions in plain language by describing your schema to a language model and generating one SELECT statement in Oracle 26ai.
How to Test AI Answers with an LLM Grader in Oracle
Check that RAG answers contain the right facts, cite real sources, and claim nothing unsupported, using SQL and a second language model.
How to Build RAG with PL/SQL in Oracle Database
Answer questions from your own data with Gemini and Oracle AI Database 26ai, citing sources and saying so when the answer is not there.
How to Extract Structured Data from Text with an LLM in Oracle
Get the details hidden in free text out with a language model in Oracle AI Database 26ai, in the right form and without invented answers.
How to Classify Data with an LLM in Oracle SQL
Sort 400 tickets into categories, priorities, and sentiment with Gemini from Oracle SQL, in batches, and measure how far to trust the labels.
How to Make LLM Calls Reliable in PL/SQL
Log every language model call, retry when the provider is busy, see the tokens you pay for, and avoid calling a model for every row of a query.
How to Call an LLM from PL/SQL with UTL_TO_GENERATE_TEXT
Send prompts to a language model from SQL and PL/SQL in Oracle AI Database 26ai and control the role, randomness, length, and format of answers.
How to Search PDF Documents by Meaning in Oracle
Find the passage that answers a question in PDF and Word documents with Oracle AI Database 26ai, and test which chunk size works best.
How to Chunk Documents for Vector Search in Oracle
Turn PDF, Word, and HTML files into searchable chunks in Oracle AI Database 26ai with UTL_TO_TEXT, UTL_TO_CHUNKS, and VECTOR_CHUNKS.
How to Build Hybrid Search in Oracle Database
Get the best of semantic and keyword search in Oracle AI Database 26ai with reciprocal rank fusion in SQL or a hybrid vector index.
How to Search by Keyword with Oracle Text
Find exact codes, versions, and product names that semantic search misses, with an Oracle Text index, CONTAINS, and SCORE in Oracle Database.
How to Build Semantic Search in Oracle Database
Find answers by meaning in Oracle AI Database 26ai, learn how near is near enough, flag unanswerable questions, and keep filters accurate.
How to Create an HNSW Vector Index in Oracle
Get millisecond vector search with an in-memory HNSW index in Oracle AI Database 26ai, sized, measured, and tuned on 50,000 rows.
How to Create an IVF Vector Index in Oracle
Speed up similarity search on large tables with an IVF vector index in Oracle AI Database 26ai, and measure what it gains and gives up.
How to Choose an Embedding Model: In-Database vs Gemini
Test embedding models on your own questions in Oracle AI Database 26ai and learn when a small in-database model is enough and when Gemini pays off.
How to Generate Embeddings with Gemini from PL/SQL
Get Gemini embeddings from SQL and PL/SQL, embed whole tables in batches, call the REST API for extra options, and survive a busy provider.
AI Applications with Oracle Database 26ai and APEX 26.1 Book
AI Applications 26ai Oracle AI Database and APEX The book Contents Examples Look inside FAQ Look insideReal pages from the book Why this book How an entry works Examples and screens Look inside What the book covers AI in the…