AI
Native FileMaker AI, embeddings, semantic search, RAG, natural-language find and SQL, and agentic model workflows.
Bring the messy part
AI making a weird face at you?
Book consulting when the lesson makes sense, but your actual file has opinions. We can untangle the design, debug the snag, or map the next build.
FileMaker AI feature map: what each step is for
Learn the current FileMaker AI script steps and functions by purpose, version, and test-worthy distinction.
Configure AI Account correctly
Use Configure AI Account with provider, endpoint, API key, and runtime scope exactly as Claris documents it.
Generate Response from Model and agentic mode
Build model-response workflows with message history, streaming, tool calls, and safe agentic mode boundaries.
Perform Find by Natural Language
Use natural language prompts to generate FileMaker find requests, found sets, or debug JSON.
Perform SQL Query by Natural Language
Ask questions of table data with generated SQL, DDL scoping, actions, and streaming behavior.
Embeddings and Perform Semantic Find
Generate embedding vectors, store them efficiently, and search text or images by meaning.
AI functions for vectors, schema, and tokens
Use GetEmbedding, CosineSimilarity, NormalizeEmbedding, GetTableDDL, GetFieldsOnLayout, and GetTokenCount in AI workflows.
RAG accounts and Perform RAG Action
Configure RAG access, add text or PDF data, send prompts, remove documents, and inspect RAG spaces.
Prompt templates and AI call logging
Use prompt templates for consistent model instructions and Set AI Call Logging for audit-friendly development.
Testing and hardening FileMaker AI workflows
Prepare AI features for production with deterministic prompts, error checks, scoped schema, and human review where needed.