fore ai
Swiss-built autonomous QA platform whose AI agents write, run and self-heal end-to-end tests for web and mobile applications. Tests are described in plain English, executed in parallel across browsers and real devices, and wired straight into CI/CD pipelines.
What is fore ai?
fore ai is an autonomous quality assurance platform for enterprise software, built in Zurich by a team of former Google engineers. Rather than asking testers to script scenarios, it puts a set of cooperating AI agents — named Explorer, Planner, Coder and Verifier — in charge of the whole cycle: understanding an application, producing test cases, running them, and repairing them when the product changes. The platform appears in its own technical documentation under the internal name Critical Journey, and browser execution runs on Playwright.
Coverage spans web and mobile on a single platform. On the web side, the stated scope runs from customer portals and transactional banking, insurance or e-commerce journeys to internal CRM, ERP and HR systems and multi-system flows with API and data dependencies. Tests execute across Chrome, Firefox, Safari and Edge and across desktop, tablet and mobile viewports. On mobile, a tester drags in an Android APK or an iOS IPA, describes the scenario in a sentence, and the generated test runs in parallel on real Android and iOS devices, returning logs, screenshots and video. Flutter, React Native, Java, Kotlin and Swift applications are all declared supported without extra setup.
Three ideas carry the product. Creation is no-code: a scenario written in plain English, or a recorded sequence, becomes an executable test. Execution is parallel and schedulable, triggered on demand, on a cron schedule, or from a CI/CD pipeline. Maintenance is automated through self-healing: when a test breaks because the interface moved, the platform proposes a repaired version alongside a diff that a human approves or rejects.
Around that core sit the pieces an enterprise buyer looks for: nine named integrations across CI/CD, issue tracking and chat, SAML single sign-on, API-key access, private network deployment, audit logging, a documented REST API and a command-line tool. The vendor also offers deployment inside the customer's own VPC or on-premises, and fine-tuning on proprietary test data. Pricing is not published; the product is sold by contract, with a demo arranged by appointment.
What it does
- Generate a complete end-to-end test from a plain-English description
- Upload an Android APK or iOS IPA and test it on real devices in parallel
- Run tests automatically on every commit through a CI/CD pipeline
- Self-heal tests broken by interface changes and approve the proposed diff
- Schedule recurring test collections and route results to Slack or Microsoft Teams
- Validate APIs, databases and modular multi-system workflows
- Drive the whole platform through a documented REST API or the fore command-line tool
When to use fore ai / When not to
A quick filter to help you decide if fore ai is the right fit.
When to use fore ai
- QA engineers and QA managers maintaining large regression suites across web and mobile
- Test automation engineers who want to retire brittle Selenium and Appium scripts
- Engineering and DevOps teams gating releases on automated checks inside CI/CD
- Regulated enterprises in banking, insurance and media that need isolated VPC or on-premises deployment
- Product and delivery managers who need release confidence without growing the QA headcount
When not to use fore ai
- Teams that need a published price list, since no pricing page or amount exists anywhere on the site
- Small teams and individuals, as the offer is sold by contract with fees payable annually in advance
- Buyers who require a named ISO 27001 or SOC 2 certificate, as only a self-declared 'Compliant' status is shown
- Organisations needing an interface or documentation in a language other than English
- Anyone looking for testing outside web and mobile applications, such as hardware or embedded validation
How to use fore ai
A typical end-to-end flow, from setup to results.
- Sign up or sign in at app.foreai.co
- Create an organisation, which acts as the top-level tenant for members and settings
- Create a test suite, normally one per application or website under test
- For a mobile target, upload the Android APK or iOS IPA as an asset
- Describe the test in plain English as a name plus step-by-step instructions
- Let the AI generation pipeline turn the description into executable steps, then review them
- Optionally save a browser state so tests start already authenticated and skip the login flow
- Group test cases into a collection and attach a cron schedule and run settings
- Connect the CI/CD platform so tests run on each build or commit and gate the release
- Route results to Email, Slack or Microsoft Teams, and subscribe to daily, weekly or monthly reports
Pros & Cons
Pros
- Web and mobile testing unified on one platform, with no scripting required
- Self-healing removes most of the maintenance burden that sinks traditional suites
- Documented public REST API with 122 endpoints, plus a command-line tool
- Nine named integrations covering CI/CD, issue tracking and team chat
- Personal data hosted exclusively in the EEA and Switzerland, with a published list of nine subprocessors
- Explicit commitment that customer data is never used to train the vendor's global models
- VPC and on-premises deployment options, with a published 99% monthly availability SLA
Cons
- No public pricing at all: no pricing page exists and no amount is published anywhere
- ISO 27001 and SOC 2 are described as a certification in the privacy policy but only as a 'Compliant' status in the vendor's own Trust Center, with no certificate number, body or validity date
- The free trial is asserted on the homepage but written as optional in the terms, which say the provider 'may offer' one
- Hosting statements conflict: the privacy policy names Microsoft Azure exclusively while the subprocessor list also includes GCP and Supabase
- The terms grant the vendor ownership of models and algorithms derived from customer data
- The product is being renamed to Klarent, so brand, documentation and application currently span two names
- Interface and documentation are English only, and the demo is by appointment rather than self-service
Pricing & Plans
fore ai publishes no price. There is no pricing page on the site, and an enumeration of the sitemap confirms none exists; no amount, currency or billing unit appears anywhere. The product is sold by contract: the terms of service state that fees are set out in an individual Order and, unless agreed otherwise, are payable annually in advance, with 5% annual interest on late payment and possible suspension beyond 30 days overdue. Free access to explore the product is advertised on the homepage, and the terms provide for an optional free, no-obligation trial period whose length is not published. Prospective buyers must contact the vendor for a quotation.
- pricing is agreed individually in a contractual Order
- Free exploratory access to the application
- advertised on the homepage
- Optional free trial period
- offered at the provider's discretion under clause 1.4 of the terms
- Enterprise engagements including dedicated integration environments and priority support under the SLA
Data, GDPR & hosting
A consolidated view of how fore ai handles your data.
GDPR overview
Implementation is concrete and documented. fore ai AG names itself the data controller and publishes security@foreai.co as the privacy contact. The policy, effective 4 May 2026, sets out purposes with their legal bases — contractual necessity, legitimate interest and legal obligation — and grants access, rectification, erasure, restriction, portability and objection rights, with valid requests answered within 30 days. Where the company acts as a processor, data subjects are directed to the customer first. Breach notification to individuals and supervisory authorities is committed to. Stated safeguards include TLS 1.2+ in transit, encryption at rest, role-based access control, least privilege and multi-factor authentication. The Trust Center lists GDPR with a Compliant status. No Article 27 EU representative is named.
Who owns the data?
The terms are explicit: the customer owns the Customer Data, and fore ai processes it on the customer's behalf under the agreement. Content captured from the applications under test is handled strictly as a processor, governed by the vendor's Data Processing Agreement. Two counterweights deserve attention. First, clause 5.3 assigns to fore ai the exclusive ownership of any data models and algorithms it derives from the customer's use and data, to exploit at its discretion. Second, the same clause allows disclosure to other customers only in aggregated and anonymised form. Each customer is isolated behind a unique Organisation ID, with scoping enforced at the API layer across the multi-tenant environment.
Reuse rights
Customers keep the right to use their own data freely: fore ai acts on instruction and claims no licence over the content itself. The vendor's own processing is bounded by declared purposes — running and improving the service, account management, diagnostics, feature development, security and fraud prevention, legal compliance, and usage analytics. The privacy policy states plainly that customer data is never used to train fore ai's global AI models, and that personal data is not sold. Sharing is limited to vetted service providers, to a merger or asset sale, and to valid legal process. The reservation to weigh is contractual rather than privacy-related: derived models and algorithms belong to the vendor, and aggregated, anonymised outputs may reach other customers.
Data retention & training
Hosting summary
fore ai AG is headquartered in Switzerland. Its privacy policy states that all personal data is stored exclusively within the EEA and Switzerland on Microsoft Azure infrastructure, and that no personal data is transferred outside that perimeter; a copy of the relevant safeguards can be requested at security@foreai.co. Customer data is logically isolated by a unique Organisation ID, with scoping enforced at the API layer in a multi-tenant environment. On termination, data is securely deleted using Azure native deletion capabilities and disposal records are kept. One inconsistency should be noted: the vendor's own Trust Center publishes nine subprocessors, and that list names GCP and Supabase under IT infrastructure in addition to Microsoft Azure. The remaining subprocessors are Slack, a GitHub application, bexio, GoDaddy, Google Workspace and Microsoft Teams. The exclusivity of the Azure arrangement should therefore be confirmed contractually.
Things to keep in mind
Risks and trade-offs to weigh before adopting fore ai.
- Pricing is entirely opaque: with no published amount and fees payable annually in advance, budget exposure cannot be assessed before entering a sales conversation
- The security posture may be overread: ISO 27001 and SOC 2 appear as a 'Compliant' self-declaration in the vendor's Trust Center, not as certificates with a number, an accredited body and a validity date, despite the privacy policy using the word certification
- Ownership asymmetry: while you keep your data, the terms give the vendor exclusive ownership of models and algorithms derived from your usage, which may matter if your test suites encode proprietary business logic
- Hosting statements conflict between the privacy policy and the subprocessor list, so any commitment about where data physically resides should be confirmed in writing before onboarding
- Automating quality assurance can erode a team's own testing judgement over time: when tests are generated and repaired by agents, fewer engineers retain a mental model of what is actually covered, and gaps become harder to notice
- Self-healing can silently mask a genuine regression by adapting a test to a change that was in fact a defect, so approving healing diffs deserves the same scrutiny as reviewing code
- Headline claims of 90% less effort, zero maintenance and 10x faster releases are published without any methodology, and should be treated as marketing rather than measurement
Setup & Integrations
Technical difficulty
Getting started is deliberately light: you sign up at the web application, describe a test in plain English, and for mobile simply drop in an APK or IPA, with no additional setup and no Selenium or Appium knowledge required. Real effort appears later and is a team matter rather than an individual one: wiring the CI/CD pipeline, connecting issue trackers and chat channels, and configuring SAML single sign-on per organisation. Integrating with the vendor's API is contractually the customer's responsibility. A VPC or on-premises deployment is an infrastructure project in its own right. White-glove support is offered for custom integrations.
Deployment
Integrations
Supported languages
Behind fore ai
Fundraising
Social
Resources
All the official URLs gathered for verification and reference.
Frequently asked questions
How much does fore ai cost?
Do I need to write Selenium or Appium scripts?
Can it test both iOS and Android applications?
What happens when the interface changes and a test breaks?
How does it fit into a CI/CD pipeline?
Is there an API?
Where is my data hosted?
Is customer data used to train the vendor's AI models?
Is fore ai ISO 27001 or SOC 2 certified?
Can it be deployed inside our own infrastructure?
Should you pick fore ai?
fore ai is a technically credible autonomous testing platform rather than a thin wrapper around a language model. The documentation is structured and specific, the REST API exposes 122 endpoints on the vendor's own backend, a command-line tool exists, and Playwright drives browser execution. The product genuinely covers web and mobile on one platform, and the self-healing loop — a repaired test presented as a diff for human approval — addresses the maintenance cost that defeats most automation projects.
Data governance is unusually well documented for a company of this size: personal data confined to the EEA and Switzerland, a nominative list of nine subprocessors, a Data Processing Agreement, a 30-day log retention period and an explicit commitment not to train global models on customer data.
Three reservations should temper that. Nothing about pricing is public, so the total cost cannot be assessed before speaking to sales, and the terms show fees payable annually in advance. The security claims are weaker than they first appear: the privacy policy speaks of ISO 27001 certification while the company's own Trust Center shows only a self-declared Compliant status, with no certificate number, accredited body or validity date. And a rebranding to Klarent is under way, so buyers will encounter the same product under two names across the website, the application and the documentation.
The natural buyer is an enterprise QA or engineering team in a regulated sector that already needs isolated deployment and can negotiate a contract. Teams wanting a self-service tool with a visible price should look elsewhere.
- Choosing a selection results in a full page refresh.
- Opens in a new window.