A custom app built on the IBM Watson and watsonx platform, for enterprises with an existing IBM stack: banks, insurers, healthcare and government organisations that want to deploy AI without losing control over data and compliance. We build on watsonx.ai, watsonx.data, watsonx.governance, Watson Assistant, Discovery, NLU and Speech-to-Text where that fits, and we are honest when another route is the better one. This is not a vendor pitch: Watson is one option alongside Anthropic, OpenAI, Mistral or a local model.
IBM Watson has quietly reinvented itself in recent years. The original Watson from the Jeopardy era has largely been wound down, and what remains under the Watson banner is a collection of enterprise AI building blocks that work exceptionally well for a particular type of organisation: those whose existing systems run on IBM (Db2, MQ, mainframe COBOL, z/OS, Power Systems, Db2 Warehouse), where a purchasing relationship with IBM is a given, and where audit, governance and data residency matter more than the very latest benchmark score.
The new centre of gravity is called watsonx and consists of three related products: watsonx.ai (foundation models and custom model deployment), watsonx.data (an open lakehouse built on Iceberg and Presto) and watsonx.governance (model oversight and AI Act documentation). Alongside these, the older but still widely used components, Watson Assistant, Watson Discovery, Watson NLU and Watson Speech-to-Text, run in the same IBM Cloud environment. We build custom apps that call these components from our own front end and our own domain logic. Our broader approach to this kind of work is set out on the AI development page, with the more general LLM layer covered on custom LLM integrations.
What sets us apart is that we are vendor-independent. We are not an IBM partner with a quota, and we have no commercial incentive to choose Watson when another route fits better. For a retail start-up with a fresh cloud budget and no IBM history, watsonx is rarely the right choice; for a Dutch bank with a mainframe in Diemen, it often is. We can talk through how that trade-off applies to your situation in a vendor advice session before we discuss building anything.
Watson projects we encounter in practice tend to fall into three patterns. Which pattern fits you depends on where the value lies: conversational interfaces for customers or employees, document AI and search over your own corpus, or a deeper custom app in which watsonx.ai forms the brain for your own domain.
An app or chat layer that handles customer or internal staff questions based on a controlled corpus. We use Watson Assistant for dialogue management where it fits, or build a custom orchestrator on watsonx.ai with your own prompts, retrieval and guardrails, depending on how much control you want over the conversation. For banks and insurers this often includes a handover mechanism to human agents and logging suitable for compliance investigations. We connect to an existing contact centre (Genesys, NICE, Avaya), CRM (Salesforce, Dynamics, Siebel) or telephony stack.
An app that reads, classifies, summarises and makes searchable large streams of documents: contracts, policies, case files, regulations, knowledge bases, process documentation. Watson Discovery serves as the retrieval layer on top of your own corpus, while watsonx.ai generates the answers, summaries or structure. For regulated industries, you can keep the corpus within IBM Cloud Frankfurt, with granular access rules per role or document type. We then build integrations with existing document management systems (FileNet, OpenText, SharePoint, NetDocuments) and mainframe sources.
A bespoke app with its own domain, its own workflow and watsonx as the engine, for example a fraud investigation tool for an insurer, an underwriting assistant for a bank, or a triage app for a hospital. We build the front end as a native or web app, handle integration with the mainframe and data warehouse, and deploy watsonx.ai for reasoning. watsonx.governance keeps track of what the model does and how that relates to the EU AI Act. Our broader approach to this kind of work is set out in enterprise AI implementation.
A production-ready custom app that draws on the Watson and watsonx building blocks where it makes sense, runs in your IBM Cloud environment or on-prem watsonx installation, and fits your existing architecture. No lock-in to us as an agency: you own the code, prompts, models and infrastructure definitions.
Eight patterns we keep seeing in our conversations about IBM Watson and watsonx. If you recognise your organisation or use case in one of them, we are happy to talk further, and we will also be honest when another platform or a hybrid approach fits better than pure Watson.
An app that helps the bank employee with credit underwriting, KYC investigation or fraud analysis, using watsonx.ai as the reasoning layer and existing core banking data from DB2 and the mainframe as the raw material. For clients under DNB and AFM supervision, audit logging and explainability weigh heavily; watsonx.governance helps build these requirements in.
An app that classifies incoming claims, detects fraud indicators, explains policy terms and proposes settlement for standard cases, while the human claims handler stays in control. Often with Watson Discovery over the policy archive and watsonx.ai for the reasoning, connected to Guidewire or a bespoke core insurance system.
Apps that turn clinical data into decision support for doctors or triage staff, for example pre-consultation review for outpatient referrals or administrative assistance with discharge documentation. For healthcare, NEN 7510 applies from day one. We deliver suitable integrations with HiX, Epic, ChipSoft and CGM, and keep audio and patient data within the EU.
Implementing organisations (tax authority, employee insurance agency, social security services, municipalities, ministries, regulators) handling large case files and dense regulation want to use AI without giving up control over data and explainability. Watson on IBM Cloud Frankfurt or on-premises Cloud Pak then fits within your existing procurement, security and architecture frameworks.
Logistics companies with a hybrid SAP/IBM architecture (DB2, MQ, mainframe batch, alongside SAP S/4HANA) are a typical profile where Watson fits well, not because it is fashionable, but because the data already lives there and the IT organisation can work with it. We build apps for procurement, planning or inventory AI that run on this type of stack.
Energy and network companies that want to build predictive maintenance, asset AI or field service apps on an existing IBM Maximo and DB2 environment. Watson Speech to Text for hands-free reporting in the field, and watsonx.ai for the reasoning around faults or asset condition. Often in line with NIS2 requirements, which are tightening for the energy and water sectors.
Supervisors and internal compliance teams working under dense regulation (Wwft, MiFID II, DORA, NIS2, AI Act) who need to find passages, precedents and risk signals. Watson Discovery on top of a proprietary corpus, watsonx.ai for summarisation and explanation, and watsonx.governance for traceability.
Organisations with a large IBM installed base (DB2 Warehouse, Db2 on z, Cognos, OpenPages) that want to take a first serious AI step without bringing in a second cloud platform. For these clients, Watson is often the practical choice; we help with the first application, the evaluation, and the decision on whether a second platform should be introduced later.
A brief orientation on each building block, as we work through them in a first session. Not everything on this list belongs in your application — we decide together which components add value.
The reasoning layer: a set of foundation models (IBM Granite, Llama, Mixtral and other open-source families) that you can use via a hosted API in IBM Cloud Frankfurt, or run yourself in a Cloud Pak in your own data centre. For clients where audio or source data must never leave the premises, the on-premises route is decisive.
An open lakehouse architecture built on Apache Iceberg and Presto, with a choice of query engine and connectivity to existing warehouses. For clients whose data is spread across Db2, Cloud Object Storage and existing analytics environments, this provides a unified layer for the AI application to rest on — without first migrating everything.
Model registration, monitoring for bias and drift, and the documentation you need for an AI Act classification as high-risk. For banks and healthcare providers that must account to DNB, DORA or NEN 7510, this is the building block that makes the difference between "we are doing something with AI" and "we can demonstrate what we are doing".
The older but still widely used conversational platform, with visual dialogue management, intents and entities. It works well for high-volume customer contact where you want tight control over flows — for example, a claims bot at an insurer. For open-domain conversations, we more often opt for a custom orchestrator on watsonx.ai.
The search and document AI layer: ingestion of PDFs, Word files, HTML and mainframe files, with semantic retrieval and entity extraction. It works well as the retriever beneath a watsonx.ai application for RAG, and it has connectors that can work with existing document stores.
Two older but still useful building blocks: NLU for categorisation, sentiment and entity extraction on text, and Speech-to-Text for transcription within IBM Cloud. For specific speech recognition questions, we also consider other engines. See the page building a speech recognition app for the broader trade-offs.
With OneDayBuild, we turn your idea into something tangible in one day for €1,150, so you can see whether further development is worth the investment. Decide to go ahead with the full build? Then we credit the full cost.
Explore OneDayBuild →A conversation in which we understand the organisation, the existing IBM relationship and the use case. We are explicit about this: not every AI application needs a Watson layer, and not every IBM client needs to put everything on Watson. Sometimes a focused advisory session is sufficient — see our page on digital consulting for the broader advisory route. If Watson is the right choice, we record why; if it isn't, we say so too. A vendor assessment belongs before the first sprint, not after it.
We map out your existing IBM stack (DB2, MQ, mainframe COBOL, FileNet, Db2 Warehouse) and agree together where the AI workload should run: IBM Cloud Frankfurt for most clients, or watsonx on Cloud Pak for Data on-premises or in a private cloud for clients with stricter requirements, such as healthcare, defence, government and some banks. We record data residency, retention, sub-processors and encryption strategy in a single architecture document, which your security architect and the Data Protection Officer can review.
For banks, insurers, healthcare providers and government bodies, a Watson app is inherently a compliance question. We carry out the AI Act classification (minimal risk, limited transparency or high risk), a first-pass DPIA and, for financial clients, the translation into DORA and NIS2 requirements. For healthcare clients, the architecture aligns with NEN 7510, as described on our page about NEN 7510-compliant software. watsonx.governance handles much of the logging requirements automatically, provided it is built into the design from the start.
A working build on a staging environment every sprint, with a growing evaluation set containing examples from your domain. Each time we swap the model, adjust the prompt or add a new RAG source, we run the evaluation set and see straight away whether quality improves or worsens for each cohort. For app development, we use the usual tooling: Flutter, React Native or native iOS/Android for the front end, FastAPI or Spring for the back end, connected to watsonx endpoints and an IBM Cloud account.
A phased rollout through a pilot group, with training for end users and a handover document for the model owner within your organisation. We link watsonx.governance dashboards to the right role, often a compliance officer or risk manager, so that oversight after go-live does not become administrative leftovers. You can have management handled by us as an agency or by your internal team. We value a clean handover and no artificial dependencies.
For organisations choosing Watson, compliance is almost always one of the reasons. We handle the legal, security and governance layer from the first sprint, not as a closing paperwork exercise, and not as an external consultancy stream running alongside the build.
For high-risk applications, the AI Act requires documentation, monitoring, transparency and human oversight. watsonx.governance is designed for this: model registration, evaluations, drift detection and risk classification sit in a single toolchain. We build that toolchain into the app, not alongside it.
For every use case involving personal data, such as customer contact, employee data, case files or claims, we carry out a first-pass DPIA and arrange data processing agreements with IBM and any sub-processors. Audio and sensitive text remain within the EU, and we show exactly which processors see what.
Banks, insurers, critical infrastructure and large service providers fall under DORA and NIS2. We align the Watson architecture, supplier management and incident response with those frameworks, so the app is not called back by risk and compliance teams at a later stage. Concretely: contractual requirements for IBM as a third party, an exit strategy, and operational resilience testing.
By default, we choose IBM Cloud Frankfurt for data residency within the EU. For clients with stricter requirements, such as some banks, healthcare providers and government bodies, we run watsonx on Cloud Pak for Data on-premises or on a dedicated private cloud. We explicitly document which route has been chosen and why.
For healthcare clients, we adapt the Watson architecture to the institution's NEN 7510 controls: access management, logging, encryption and supplier management. We align with the existing ISMS and with the working practices of the medical ethics committee (FG) and the information security function.
Every interaction with the Watson or watsonx system leaves a traceable record: which model, which prompt, which retrievals, which output and which human check. For high-risk applications, we build explicit "human-in-the-loop" steps in before the output goes live.
What clients typically want to know before we start a Watson or watsonx project.
A no-obligation half-hour introductory call. Tell us who will use the app, in which architecture and with which compliance context. We'll think along, give direction, and be honest about when watsonx is the right choice and when another platform or a hybrid approach fits better. Sometimes a short advisory phase is the right first step before we talk about building. You can also email directly at fabian.vandijk@appfront.nl.
Appfront uses cookies and similar technologies to keep the website working properly, for analytics and for marketing. You choose what you allow. Read more in our privacy policy.
