FileMaker's native AI features are just the tip of the iceberg. In this session, we'll go beyond the platform's built-in capabilities and explore how to call external AI APIs directly from FileMaker using Insert from URL — no middleware, no external servers. The focus is on multimodal generative models: how to send images, audio or any other kind of file to frontier AI providers (OpenAI, Anthropic, Google) straight from your scripts, handle Base64-encoded content, structure JSON payloads, and get actionable results back into your database. A practical, hands-on session with real deployable patterns you can take home and use immediately.
This session explores three areas I’ve been spending a lot of time with recently: measuring performance, managing technical debt, and AI-assisted development. While they may seem unrelated at first, they all connect back to something FileMaker has enabled for decades — rapid iteration and the ability to continually evolve what we build.
Today, AI-assisted workflows and “vibe coding” are accelerating that capability even further, creating both exciting opportunities and important new challenges.
Part practical session and part open discussion, I’ll share experiments, observations, and lessons learned from current projects, while also exploring the broader implications for our community, our workflows, and the future of software development itself.
How do small businesses prepare for mandatory Peppol e-invoicing? In this session, I’ll share practical experience integrating Peppol into FileMaker applications using the Billit API, including architecture choices, validation challenges, client onboarding, and implementation lessons learned.
OData Webhooks introduce an event-driven paradigm into FileMaker, enabling your solutions to react to data changes in real time. In this session, we’ll explore how to design and implement event-driven workflows using native capabilities, with a focus on simplicity, reliability, and practical use cases.
Managing FileMaker Server schedules can quickly become complex, especially when dozens of active tasks are already running on the server.
What happens when a new schedule needs to be created, or an existing one must be modified? In many cases, we first need to analyze how the new task might affect the current scheduling structure. With a large number of existing schedules, this often turns into a time-consuming and error-prone process.
But what if there were a graphical interface that visualized all schedules in a clear and intuitive way? A visual overview could make it significantly easier to understand dependencies, identify conflicts, and optimize timing at a glance.
Because, as the saying goes: "A picture is worth a thousand words."
Here is the link to the repo with more information: https://github.com/agametis/fms-schedule-visualizer
Modern Claris solutions rarely live in isolation. Today’s applications must connect to payment systems, messaging platforms, AI services, accounting software, and partner ecosystems.
In this session, we will explore modern integration strategies for FileMaker and Claris solutions — from low-code workflows with Claris Connect and webhooks to scalable API-first architectures.
Using real-world examples from production environments handling tens of thousands of daily workflows and more than 150,000 API calls per day, this session covers:
Claris Connect workflows
Webhooks and automation
The transition from Data API to OData
Centralized API architecture
Scaling integrations with redundancy and load balancing
Building reusable integration layers
Attendees will learn how to design integrations that are simpler to maintain, easier to scale, and ready for the future.
Do not rely on AI or vibe coding for your reporting needs!
Reporting in FileMaker can feel limiting as requirements grow. What many teams don’t realise is that a lot of enterprise customers already have access to Power BI through Microsoft 365 — it’s simply sitting there, unused.
That makes Power BI a surprisingly accessible option for serious reporting. In this session, I’ll show how it can complement FileMaker, not replace it — and how to use it properly.
I will show you:
How to prepare your FileMaker database for PowerBI reporting
Setting up the foundations in PowerBI
Some cool visualizations
A couple of technical things that took me a while to figure out, like date tables.
The licensing model: creating dashboards versus sharing…
Why this is relevant for us as Claris Platform consultants.
On the topic of getting AI generated code (scripts et al) into FileMaker there are two camps:
backdoor people, who purport the use of the FileMaker Update Tool to patch code into solutions
frontdoor people, who use the FileMaker Client GUI to 'do work as normal' and (mostly) use the FileMaker clipboard (fmxmlsnippet et al) to grab & patch code
Whilst API-based AI-interaction is, of course, the automation dream we strive for (and odata already provides great possibilities here) the FileMaker Update Tool is still very wanting in a multi-developer server environment (dare I say useless for now?), and I am thus very much (at least for the time being) in the frontdoor camp.
To boot I have a host of tools for enhancing the FileMaker GUI for increased productivity here (see fmWorkMate.com and one of them has an important role to play in the new era of AI…
fmIDE - a FileMaker Integrated Development Environment - is an integration tool (a one-script (*) module for each of your FileMaker files) which opens up a bridge between your AI agent and the FileMaker GUI.
With the fmIDE 'Name that Thing' API you / your AI can jump to any (**) thing in FileMaker in a single click (***).
With fmIDE Action Scripts - and the newly emerging and very exciting fmJAML format - you and your AI are also able to write dynamic action scripts that can not only select steps in scripts, but also replace the code.
fmIDE is already integrated into BaseElements, now also into Marcel Moire's coming AI tool, … and could soon be on your desktop.
Approximate content:
- State of AI FM-Code generation / FM-Challenges / step text, xml & co. / getting code back into FM: frontdoor/backdoor camps
- A short (re-)introduction to fmIDE, what it does, where it fits in to the modern AI supported code generation: as an action bridge between AI and FM.
- fmIDE Name that Thing API
- fmIDE action scripts
- Diagramme of how fmIDE works inside
- JSON as a structured data language - limitations ('fluffy')
- JSON in FileMaker: JSONSetElement pros and cons
- fmJAML "FileMaker JaSON Markdown Language" (short presentation)
- hardening fmIDE for real world use: Proof by Testing (thanks to fmJAML)
- (frontdoor) Actions for AI / Rename Script / Update Script / etc.
Your clients found the building. Great. Now they're walking in circles looking for the entrance. Not so great.
In this session you'll learn how to integrate Mapbox interactive maps directly into FileMaker using the Web Viewer with real-time routing, custom markers and the ability to guide your users to the exact entrance of a building. Not just the address. The actual door.
We'll cover everything: dynamic placeholders, transport modes, color customization, and that satisfying moment when the map draws a dotted line straight to the front door.
Oh, and one more thing: this is an AI-free session. No LLMs. No embeddings. Just FileMaker, some HTML, and maps that actually tell people where to go.
We are using Claude-Code with Marcel Moré's fm-lab (private-beta) to research databases, develop more securely and support the developers in not missing a dependency.
I will talk you through the decisions, on how to set this up and how i did it, show you the perks of Marcel's system and how we manage a team of developers right now and how we plan to do this in 6 months.
Are you tired of spending your money on API request? Learn another way to connect your Filemaker apps with services in real time using Webhooks and OData. We are going to see how they work from end-to-end
A short introduction in the new AI server & some impressions from beta testing it. Following discussion about the concept and the product itself. (Alexander Feder & Michael Heider)
I've been teaching free AI Classes every week, and building a large set of AI design patterns and features that provide immediate value to your databases. These are all copy-and-paste soutions, in production and not just theoretical.
Incredibly powerful search across your whole db (This is not the simple Semantic search you've seen before)
Reports: What works, what doesn't, and why
Creating consistent, customized Image descriptions, search, and getting data from URLs
PDF processing, OCR and analysis
Identifying the key info users need to know when viewing complex records.
Where I’m going next: OpenClaw, auto-healing of data, and why FileMaker is relevant, and will be going forward
A demo file will be provided. All live interaction, < 5 slides.
A short practical session about FileMaker indexing: value index vs. word index, the meaning of “Minimal”, automatic index creation, and a few surprising search behaviours with separators like hyphens.
I’d love to host an open round‑table on what works (and what clearly doesn’t) when running multiple projects with a team.
Do you work with fixed weekdays per client? Have you switched to multi‑day sprints? Something else entirely? What are the real pros and cons you’ve experienced?
I’ll briefly share how we used to work at ClickWorks and what I learned the hard way about context switching. Mostly, I’m hoping to learn from your experiences and maybe even challenge a few assumptions together.
A Linux / Ubuntu based FileMaker server sometimes needs to access data on a Windows file server. This is possible with the Windows NTLM username/password authentication. But NTLM is insecure and many companies do not tolerate it in their networks anymore.
I would like to show how to access files on a Windows server using Kerberos ticket based authentication and autofs. You will need to know your way around the Linux command line and have access to a Windows domain controller.
Bring your imagination to this interactive session for forward-thinking FileMaker developers. I’ll showcase my aging application for a newspaper and digital print company, and open the floor to creative ideas for transforming it with AI integrations, modern UI/UX, richer visual features, smarter workflows, and any bold concepts that come to mind.
Whether you enjoy brainstorming next-generation solutions, reimagining legacy software, or just maybe want a laugh, this session is for you.. Expect live discussion, spontaneous ideas, and who knows? Maybe we'll build it out together.
This thing always comes up once the FileMaker solution has been finalized. I want to name important information about Claris licenses. Maybe this could start a discussion and créâtes a wishlist for future, esaier to use pricelists?
Licensing Models
Choosing the Best Price Tier
The Wild Card: Site Licenses
Using an FDS
But nothing about SBA!
This thing always comes up once the File...
Day 2
15:00-15:45
Mezzanin - Breakout (14 seats)
Klemens Kegebein
Day 2 | 15:00-15:45 | Mezzanin - Breakout (14 seats)
I’ve thrown together a set of AI claims that I keep hearing, some smart, some lazy, some half-true, and I want to test them with the room. I’ll reveal each claim one by one, then we can argue about it, banter a bit, and see which ones actually hold up in the real world. ... I have opinions too of course ;)
This is about getting alerts from the application that may och may not be hidden errors using Sentry and fixing them before your customer reports a problem.
As we develop on multiple servers for multiple files we wanted to have one dedicated InspectorPro server that collects all the data. It also shields the DEVservers from the load the primary analysis takes and therefor makes for a smoother development experience.
I will show the tweeks we made to make this work for us -- and am looking forward to a discussion on how to further improve.
As we all know, you can use multiple related table occurrences in your query to FileMaker OData. But there are some caveats the documentation doesn't explain well. In one of our projects we had to deal with those and I would like to share our findings with you.
I've always been a VR enthusiast but always considered the price tag too high to justify buying a headset to my wife. About a year ago ClickWorks gave me the opportunity to acquire one for my professional needs.
It became Meta Quest 3, I've been using it in the office hours full on since and I want to share my experiences.
We will leave together from MuseumsQuartier at 18:30 and travel via tram to the Heuriger Sissi-Huber
Roterdstraße 5 1160 Wien Österreich
sisi-huber.at
There we have ordered a buffet and an initial round of local wine, so that nobody needs to go hungry. If the weather permits, we will sit outside but in case of rain we have a cozy room to draw back to.
Please ensure that you are registered for this event, as additional costs apply.
We will leave together from MuseumsQuart...
Day 2
18:30-00:00
Social Event Heurigen Sissi-Huber
Philipp Puls
Day 2 | 18:30-00:00 | Social Event Heurigen Sissi-Huber
Informal panel discussion about the approaches to AI for the FileMaker dev, how and why it’s critical to stay current (at whatever level), and what the experts think. Pros, cons, observations, predictions. Why did the world change fundamentally in November 2025.
Let's demystify fmIDE and see how easy it is to get fmIDE into our solution and what it can do for us.
We'll first build fmIDE into one of Phil's solutions, then see how we can use Marcel More's fm-lab environment (which already supports (a subset) of fmIDE integration) to get directly to the code.
Maybe we can implement some dynamic fmJAML actions to give AI the ability to control the FM GUI too.
Not a presentation. Just an open discussion where ETS members and Claris people (except Marie who's "just an observer") can answer your questions and exchange ideas. Depending on the duration of the discussion, we'll switch to Jan Zelenka's presentation on ClickWorks Template in the same room.
Let's dicuss on how we all will be billing our customs in a years time.
What are the expectable pushbacks from customers
Value-based pricing
how to transition
This is no presentation, it is an open discussion
--
We're talking about how we argue our prices in the FileMaker development world, when vibe coders can produce 'the same' in a much shorter time. What arguments do we have, to convince customers to stay with us, compared to the attractive looking 'quick-win' of vibe coded appa. short bullet points thanks
Thought for a couple of seconds
“Same screen” is not the same system. A vibe-coded app can mimic the visible UI; it rarely reproduces the business rules, edge cases, permissions, audit trails, imports, exports, invoices, reports, backups, and support workflows.
FileMaker systems are usually living business infrastructure. The cost is not building version 1 quickly; the cost is keeping version 37 reliable after years of customer wishes, legal changes, staff turnover, and data growth.
We know the hidden domain logic. The customer often cannot fully specify what they do. Experienced developers extract, challenge, and stabilise the business process before turning it into software.
Fast prototypes create false confidence. A demo that works with five test records may collapse with real data, multi-user access, permissions, printing, integrations, and exceptions.
Cheap code becomes expensive maintenance. The question is not “Can someone generate it?” but “Who understands it, fixes it, documents it, migrates it, and is accountable when payroll/invoices/production stop?”
Data is the real asset. We protect data structure, migration paths, validation, backups, recoverability, and long-term ownership. Pretty generated apps often underestimate this.
Security and permissions are not decoration. Business apps need role-based access, auditability, GDPR awareness, safe deployment, and protection against accidental damage.
Integration is where reality bites. E-invoices, DATEV, ERP, SQL, APIs, PDF output, barcode scanners, email, Excel, web viewers, server schedules — these need durable engineering, not just UI generation.
We reduce business risk. The customer is not paying for keystrokes; they are paying for fewer failures, fewer surprises, fewer broken processes, and someone responsible.
We build updateable systems. A quick app is attractive until the next change. Professional FileMaker work includes naming, structure, conventions, deployment, testing, and upgrade paths.
AI makes good developers faster, not unnecessary. The value shifts from typing code to judging code, designing architecture, finding edge cases, and knowing what must not be automated blindly.
The customer does not want an app; they want an outcome. Faster invoicing, fewer errors, better control, happier staff, legal compliance, smoother operations. Price should be argued against that value, not against hours.
Ask the key question: “When this becomes critical to your business, who do you want responsible for it — a quick-win generator, or someone who understands your business and will still be there when it breaks?”
As most FileMaker professional, we at ClickWorks have a template we use for creating new FileMaker files. I want to present the key priciples coded into the template and have a conversation on what you do or would do differently.
Have you ever dreamed of being able to search within FileMaker scripts and find everything they contain, with operators, filters, and everything else you could wish for? In this session, I’ll show what it is possible right now.