Online guide
Guide
How to Build a FileMaker MCP Server
Connect Claude and other AI agents to your FileMaker data: the no-code route, the custom-server route and the prerequisites.
12 min read · Updated July 2026
AI & LLMs Inside FileMaker - What Actually Works in 2026
Past the hype, past the demo videos: the AI use cases that are live in real UK FileMaker deployments in 2026, which LLM to pick, UK GDPR implications, and what it actually costs.
Updated April 2026 · 11 min read · By Neptune Digital
Every client conversation in 2026 eventually turns to "can we add AI to our FileMaker system?" The answer is yes, it's easier than it was two years ago, and most of the demos you've seen online are technically accurate but commercially misleading. This guide is the honest version - the use cases that work in production, the ones that don't yet, what we deploy, what we charge, and the UK GDPR considerations you need to get right.
What Claris ships natively in 2026
Claris has invested meaningfully in AI-in-FileMaker since 2024. The native tooling as of the current FileMaker 2025+ releases:
- AI Model script step - configure a provider (OpenAI, Claude, Gemini, Ollama, Azure) once; call it from any script with a prompt + input.
- Embedding script steps - generate vector embeddings for any field or text input without leaving FileMaker.
- Vector search / semantic search - built-in semantic search against embedded fields, with a cosine-similarity score.
- Function calling / tool use - models can call scripts and return structured results for multi-step workflows.
- Privileged access controls - AI feature use can be gated per privilege set, so users with no AI budget don't rack up calls.
On top of this, any script can still call any LLM's REST API directly via Insert from URL - so if you need a model or feature Claris hasn't wrapped, you're not blocked.
Five AI use cases that actually work in production
1. Semantic search
This is the clearest win we see. Legal, clinical, CRM and knowledge-base use cases where users search across long-form notes fields ("find cases like this one") move from frustrating-literal search to genuinely useful. Typical delivery: 2–4 weeks on an existing system, £4,000–£10,000.
2. Document summarisation & extraction
Drop a PDF (contract, CV, inspection report, invoice) into FileMaker; an LLM extracts structured fields (counterparty, renewal date, total value, risk flags) with a confidence score. Human verifies. We see this live in procurement, legal, HR and insurance workflows. Typical delivery: 3–6 weeks, £8,000–£20,000.
3. Classification & routing
Inbound enquiries, tickets, referrals, leads - an LLM classifies them (sentiment, priority, routing team, likely resolution category) and updates the record. Cheap, high-volume, usually well-bounded enough to trust in production. Typical delivery: 2–3 weeks, £4,000–£10,000.
4. Draft generation (with human-in-the-loop)
First-draft quotes, response emails, case notes, executive summaries, meeting minutes. The operator always approves before send. This works; what doesn't work is removing the human from the loop - the quality and hallucination risk isn't there yet for autonomous send. Typical delivery: 2–4 weeks per workflow, £3,000–£8,000.
5. RAG-style Q&A over your FileMaker data
Chat interface inside FileMaker that answers natural-language questions grounded in a scoped subset of your data ("which suppliers missed SLA in Q1?", "what's open on the Millbrook account?"). We build these with embedded records + tool-use so the LLM cites the records it used. Typical delivery: 4–8 weeks, £12,000–£35,000.
What doesn't yet work reliably
- Fully autonomous agents that act on your database without human review. Too much hallucination risk, and the blast radius of a wrong write is high.
- High-stakes factual accuracy - compliance reports, medical advice, legal opinions - should not be LLM-authored without expert review.
- Long-context, multi-step orchestration - reliability drops sharply past 8–10 agentic hops in current models.
- Numerical analysis at scale - LLMs are poor calculators. Use Claris's SQL / aggregate functions for the maths; LLMs for the language layer on top.
Which LLM should you use?
| Provider | Strengths | Best for | UK residency option |
|---|---|---|---|
| OpenAI (GPT-5) | Generalist quality, broad tool use, fast iteration | Most non-regulated use cases | Via Azure OpenAI (UK regions) |
| Anthropic Claude 4.x | Long-context, instruction-following, safety | Long documents, sensitive tone | AWS Bedrock / Anthropic enterprise |
| Google Gemini 2.x | Multimodal (images, docs, video) | Document understanding, OCR | Via Google Cloud UK regions |
| Azure OpenAI | UK residency, enterprise compliance, MS DPA | Regulated UK deployments | Yes, native UK regions |
| Self-hosted (Llama, Mistral via Ollama) | Cost, control, true UK-only data | High-volume, regulated, offline | Yes - you own the infrastructure |
AI + UK GDPR - the things you must get right
- DPA with the provider. No DPA, no sending personal data. Full stop.
- Zero-retention enterprise tier. Consumer ChatGPT / Claude accounts do log prompts for training; enterprise API tiers don't. Always use the enterprise tier.
- Lawful basis documented. Updating your ROPA (Record of Processing Activities) for AI processing is not optional.
- DPIA for special-category data. Health, legal, criminal, biometric → expect to do a DPIA.
- DSAR coverage. Log what went where; be able to answer "you held my data, who else did you send it to?".
- UK residency where required. NHS, regulated finance, some public-sector → Azure OpenAI (UK) or self-hosted.
- Privilege-gated AI features. Not every user should be able to rack up £££ of API calls.
For the broader compliance picture, see FileMaker Security & UK GDPR - 2026 checklist.
What a good AI-in-FileMaker deployment looks like
- One scoped, well-bounded use case delivered first (not "let's AI the whole system").
- Provider chosen with compliance in mind; DPA signed and on file.
- Prompt templates version-controlled in FileMaker (not pasted into scripts).
- Every AI call logged - input, output, model, cost, user, timestamp.
- Confidence thresholds enforced - low-confidence outputs go to a human queue, not into the record.
- Cost dashboard so the business sees what AI is costing, per use case, per month.
- Kill-switch: a privilege-gated flag to disable the AI layer globally if a provider has an outage or a compliance question.
The short answer
Adding AI to a FileMaker system in 2026 is a solved engineering problem for bounded, scoped use cases. Start with semantic search, document summarisation or classification - each delivers clear ROI in 2–4 weeks. Pick a provider with a UK-compatible compliance story. Log everything. Keep a human in the loop on anything high-stakes. Don't try to autonomise your whole business in one go - nobody is doing that successfully yet.
Ready to let an AI assistant query your data directly? See how to build a FileMaker MCP server for the no-code and custom routes.
See also - FileMaker Security & UK GDPR, Claris Connect vs n8n vs Zapier, and Is FileMaker still worth using in 2026?
AI & LLMs in FileMaker FAQs
Yes - and much more easily than in 2023. Claris has added native AI features to FileMaker including an AI Model script step, semantic search with built-in embeddings, and direct integration with OpenAI, Anthropic, Google and local models. On top of that, FileMaker's Insert from URL script step can call any LLM REST API directly (OpenAI, Anthropic Claude, Google Gemini, Mistral, open-weights via Ollama, Azure OpenAI for UK-hosted). In 2026, adding AI to a FileMaker system is a days-to-weeks project, not a research project.
Five use cases work reliably in 2026: (1) semantic search across unstructured fields (notes, case history, PDFs); (2) document summarisation - turn a 40-page PDF into structured fields; (3) intelligent classification - route inbound enquiries, tag records, score leads; (4) draft generation - first-draft emails, quotes, report narratives that a human approves; (5) RAG-style Q&A over a subset of your FileMaker data. What doesn't yet work reliably: fully autonomous agents acting on your database without human review, and anything requiring 100% factual accuracy over long prompts.
For most UK use cases, either OpenAI (GPT-5) or Anthropic Claude via direct API, or Azure OpenAI if you need UK/EU data residency and enterprise compliance. Google Gemini is strong for image + document understanding. For cost-sensitive high-volume tasks, open-weights models (Llama, Mistral) run on a local/self-hosted Ollama server can be a fraction of the price. Claris's own AI Model script step abstracts most of this - you configure the provider once and switch without rewriting scripts.
Depends on the provider, the data, and your contracts. OpenAI, Anthropic and Google all offer zero-retention enterprise tiers where API data is not used for training and is not logged beyond the API call. Azure OpenAI runs in Microsoft's tenancy (UK regions available) under Microsoft's DPA. For regulated UK data (healthcare, legal, public sector) we usually recommend Azure OpenAI (UK-hosted) or a self-hosted Llama/Mistral on UK infrastructure rather than direct calls to consumer LLM APIs. For non-regulated business data, enterprise-tier direct calls are normally fine - with a DPA in place.
Normal FileMaker search is literal - you search for 'contract' and you get records containing 'contract'. Semantic search embeds your fields as vectors and returns records that are conceptually similar - so a search for 'contract' surfaces records about 'agreement', 'terms', 'MSA', 'SOW' too. Claris FileMaker 2024+ ships a native Embed and Vector Search set of script steps; you can also self-host with pgvector or a dedicated vector database. It's transformative for case management, legal, clinical notes, and knowledge bases.
Partially. GPT-5 and Claude 4.x are decent at FileMaker script logic, formula syntax, and custom function calls - they will save a developer time on boilerplate and first drafts. They still make mistakes on version-specific syntax, privileges, layout object references, and script triggers in context. In practice, we use LLMs as pair programmers - generate a draft, verify against the actual schema, test - not as autonomous developers. They are not a substitute for a Claris Certified engineer, but they meaningfully boost productivity for people who already know what they're doing.
Two costs: the build (our side) and the LLM usage (your side). Build costs typically range £4,000–£15,000 for a focused AI feature - semantic search over a table, a summarisation workflow, or a classification pipeline - and £20,000–£60,000 for a fuller rebuild of a module around AI. LLM usage depends on volume and model; typical enterprise FileMaker clients running 5,000–50,000 AI calls a month spend £40–£400/month on API usage. Azure OpenAI and self-hosted open-weights models can cut that further.
The AI layer is a sub-processor of any personal data you send it. You need a DPA with the LLM provider, a lawful basis documented for the AI processing, and - if you're sending special-category data - a DPIA. Data Subject Access Requests remain your responsibility to fulfil even when data has been passed through an LLM, so you should log what was sent, to which model, and when. See our FileMaker Security & UK GDPR guide for the broader checklist.
