AI in your applications

AI features, right where your people work.

FileMaker now ships AI as part of the platform: semantic search, AI script steps and functions, connections to language models. From these, we build features your users work with directly in the solution they already know - no new software, no tool switching.

Semantic search
Find what is meant.

Users search in natural language instead of exact field values. The search understands meaning: a query for «complaints about delivery delays» also finds the entry that says «goods arrived three weeks late» - without anyone having to guess the right keyword.

This works across records, notes, and attached documents, within the existing access privileges.

Client files
Complaints about delivery delays in spring
Meier AG - Ticket #2411
«Goods arrived three weeks late …»
Bühler GmbH - Note from Apr 14
«Client upset, appointment postponed twice …»
Keller & Co - Email attachment
«… delivery outstanding, please advise»
Results ranked by meaning, not by keyword
Questions to your own data
Questions in plain language.

An assistant right inside the solution answers questions from your records: «Which orders are critical this week?» instead of five filters and three layouts. The answer references the records it came from - one click, and you are there.

If the system finds nothing, it says so. It does not invent an answer.

ASSISTANT · ORDER MANAGEMENT
Which orders are critical this week?
Three orders need attention:
A-2381 Rüegg AG - delivery due Friday, material not expected until Thursday
A-2374 Baden Hospital - approval outstanding for 6 days
A-2369 Wyss Bau - budget 94% used
Open A-2381Open A-2374
Classify & tag
Sorted before anyone looks.

Incoming feedback, emails, or documents are categorized, tagged, and prioritized on arrival. What's urgent sits on top; similar items are grouped. Your people start the day with a sorted list instead of a full inbox.

You define the categories - the solution does the assignment, correctable with one click.

Inbox · todayautomatically categorized
ComplaintWrong quantity delivered, client waitingHIGH
InquiryQuote requested for service extensionMEDIUM
Praise«The installation team was outstanding»LOW
InvoiceSupplier invoice, amount extracted: CHF 4,820.-MEDIUM
Image recognition
Photos become searchable too.

Incoming goods, inventory, construction-site documentation: one photo with an iPhone, and the solution suggests a category and assignment. The image library becomes semantically searchable - «damaged packaging» finds the matching photos, without anyone ever having entered a keyword.

damaged packaging
98%
91%
84%
New photo → suggestion: category «Incoming goods», delivery #883
Anomalies & forecasts
See outliers before they get expensive.

Your solution knows what normal looks like - order volumes, throughput times, consumption. AI flags what deviates from it and estimates where a series of numbers is heading. Right in the dashboard your people open anyway.

THROUGHPUT TIME · DAYS PER ORDER
Outlier Forecast
In practice
How businesses use this today.
Incoming goods with photo checks

A plant retailer documents deliveries with iPhone photos. AI assigns each image to an inventory category and helps decide whether a delivery is accepted or rejected - incoming inspection became noticeably faster.

Prioritizing job applications

An HR solution compares incoming applications with the job profile, ranks the best-fitting candidates on top - and recognizes when someone is a better fit for a different open position.

Analyzing customer feedback

Product reviews are automatically grouped and ordered by need for action. Customer service starts with a finished report instead of hundreds of unread responses.

Examples from the FileMaker ecosystem, documented by Claris (claris.com).

How we work
AI is also part of our own development process.

Agentic development, prototypes from conversations, automated tests - how we build with AI.

See AI Development
Find out how AI can extend your existing FileMaker applications!
Free initial consultation