GFoundry
GFoundry is a Portuguese talent management platform that layers gamification and AI agents over HR processes, bringing learning, performance reviews, OKRs, recognition, surveys and internal communication together in one mobile-first, EU-hosted workspace.
What is GFoundry?
GFoundry is a business-to-business talent management platform published by the Portuguese company GFoundry Lda, in operation since 2014. Rather than replacing an administrative HR suite, it positions itself as a talent activation layer that sits on top of systems such as SAP SuccessFactors or Workday and adds daily engagement to them.
The platform is modular. Twelve building blocks switch on or off as needed: a full LMS (Learn & Info), the Engagement Thermometer for pulse surveys, a Marketplace where virtual coins buy vouchers, days off, experiences or charitable donations, plus Comparisons, KPI Competitions, Evaluation & Careers, Recognition, Innovation, Surveys, Goals (OKR), Tasks and Feedback. All of them rest on a shared foundation: a multi-tenant, multi-container architecture in which a head office and several business units share one account under their own branding; a Web App alongside native iOS and Android clients hosted on Google Cloud Platform; single sign-on through SAML, Active Directory, LDAP, Google or LinkedIn; role-based segmentation; and an open API.
Gamification is native rather than bolted on. Leaderboards, badges, virtual coins and Missions — visual journeys made of steps and goals, each step able to call any module — run through onboarding, upskilling, culture programmes and recognition schemes alike.
The AI layer is called GFoundry Intelligence, or Gi. Gi is a conversational assistant grounded in the documents, policies and content each organisation defines; it answers in 26 languages, respects each person's access rights and cites its source every time. Four specialist agents work around it. Gi Learn, the Learning Designer, turns a PDF or a briefing into a publishable course. Gi Practice, the Role-Play Coach, rehearses difficult conversations out loud, reacting to tone and pace before returning structured feedback. Gi Talent, the Talent Strategist, answers people-analytics questions in natural language, reads the organisation through nine behavioural archetypes and produces alerts, CHRO narratives and coaching scripts. Gi Assist, the Workplace Assistant, routes everyday employee questions to the right specialist.
Three integration routes are offered: no-code middleware through Make.com or Zapier, a REST API with webhooks for SAP or Microsoft Dynamics 365, and a read-only MCP server exposing talent data to Copilot, ChatGPT or Claude.
What it does
- Generate a complete course — objectives, modules, quizzes and resources — from a PDF or a short briefing
- Run performance and potential cycles with four evaluation models, a 9-box matrix, exit-risk scoring and individual development plans
- Query talent data in natural language and export the answers to Excel, with behavioural archetypes, alerts and per-manager coaching scripts
- Answer employees' everyday questions from the company's own knowledge base, with the source cited on every answer
- Measure engagement continuously through pulse surveys, eNPS and nine tracked metrics drawn from a library of 130+ ready-made questions
- Turn external CRM, ERP, helpdesk or BI metrics into individual and team competitions with badges and virtual coins
- Drive the platform from Microsoft Copilot, ChatGPT or Claude through a read-only MCP server
When to use GFoundry / When not to
A quick filter to help you decide if GFoundry is the right fit.
When to use GFoundry
- HR and people teams in mid-sized to large organisations, from a few hundred employees upwards, that already run an administrative HR system and want an engagement layer on top of it
- Multinational groups and holdings with several brands or business units, which the multi-container architecture keeps separate while consolidating data for the parent organisation
- Learning and development teams that need a full LMS with certificates, curricula and AI-generated course material built from their own PDFs and briefings
- Frontline and deskless workforces in retail, logistics, manufacturing or field services, reached through branded native mobile apps rather than a desktop
- Sales, call-centre and operations managers who want to turn CRM, ERP or helpdesk metrics into team and individual competitions
When not to use GFoundry
- Small companies and teams beyond five people: the free Base plan stops there, and the Plus and Premium plans start at 250 licensed users sold in blocks of 250
- Organisations looking for administrative HR software: payroll, absence approval, time clocking and contract signature are outside the scope, and the vendor's own comparison page points to Factorial or cegid PHC for those
- Buyers who want a single capability: modules cannot be bought individually, only as one of three packaged subscription plans
- Teams that need month-to-month flexibility, since Plus and Premium follow a twelve-month licensing cycle with no monthly payment option
- Developers looking for a self-serve, documented API to build on: the REST API, webhooks and MCP server are described commercially but no public technical reference is published
How to use GFoundry
A typical end-to-end flow, from setup to results.
- Take the online assessment, or run the HR ROI calculator, to check whether the platform fits the organisation
- For an immediate start, create a self-service Base account from the sign-up page; for an enterprise project, book a demo through the form
- Choose a plan: Base for self-service, or Plus and Premium from 250 users upwards, licensed in blocks of 250
- Switch on the modules that serve the current cycle — most customers activate them progressively across each twelve-month cycle
- Configure the visual identity of the apps and, where several business units exist, set up the multi-container structure
- Connect authentication: SAML, Active Directory, LDAP, Google or LinkedIn single sign-on, or standard registration by email or employee number
- Wire up the existing stack — Make.com or Zapier for no-code automations, the REST API and webhooks for the ERP or HRIS, the MCP server for AI assistants
- Load the organisation's documents, policies and content so that Gi can answer from them and the agents can generate material
- Build the Missions that structure the journeys, then launch a first focused use case such as onboarding or a training academy
- Expand in waves with a certified partner: full transformation programmes are described as running across twelve months
Pros & Cons
Pros
- Unusually broad functional scope for one product: twelve modules spanning recruitment, onboarding, learning, performance, recognition, innovation and internal communication
- Explicit privacy posture on the AI layer — tokens instead of names, server-side identity resolution, tenant isolation, and no training on customer data
- Data processed and stored in the European Union on Google Cloud, with Standard Contractual Clauses and the Data Privacy Framework for transfers
- AI subprocessors named openly, OpenAI OpCo LLC and Google's paid-tier Gemini, with the substance of both Data Processing Agreements published
- Three genuine integration routes rather than one: no-code middleware, REST API with webhooks, and a read-only MCP server for AI assistants
- Prices published openly, which is rare in talent management, and a permanent free plan for up to five users that allows real evaluation before any commitment
- Documented customer references with named case studies across logistics, retail, banking, pharmaceuticals and consulting, plus a substantial public help centre
Cons
- High entry threshold on the enterprise plans: a 250-user minimum, licensed in blocks of 250, with a twelve-month cycle and no monthly payment
- Modules cannot be bought individually — only three packaged plans are sold
- Published prices are stated as an upper bound at the 250-user tier and decrease with volume, so the real cost cannot be worked out from the site
- The pricing page is rendered by a third-party widget: the detailed comparison tables do not display without JavaScript, and some still carry the widget's demonstration amounts
- No public API documentation: the REST API, webhooks and MCP server are described commercially but no technical reference is available to evaluate them
- No customer-facing Data Processing Agreement is published or explicitly offered, and the SOC 1/2/3 reports and C5 certification listed on the governance page belong to Google Cloud, not to GFoundry itself
- Several published details are inconsistent — the registered address differs between the privacy policy and the footer, the number of supported languages is given as 26 on some pages and more than 30 on another, and the Terms of Service still date from July 2023
Pricing & Plans
A permanent free plan is available: the self-service Base plan is entirely free for up to five users and opens a limited set of modules. The lowest paid entry point is the Plus plan, advertised at up to 9.00 EUR per user and per month. Premium is advertised at up to 13.00 EUR per user and per month. Both figures are stated as ceilings applying at the first licensing tier of 250 users; GFoundry licenses in increments of 250 users and the unit price decreases as the number of licensed users rises, so the effective cost must be quoted. Plus and Premium include consulting and integration hours from a certified partner, require a setup fee, and follow a twelve-month licensing cycle. Monthly payment is available on the Base plan only, and fees are non-refundable except where expressly stated.
- Base — self-service
- from 0.00 EUR per month and free for up to five users. Opens Learn (Lite) micro-learning
- OKR Management
- Feedback
- the Engagement Thermometer for pulse and insights surveys
- the Marketplace and the gamification engine. Text support included
- monthly payment available
- no setup fee
- Plus — enterprise plan
- up to 9.00 EUR per user and per month at the 250-user tier
- flagged as the popular choice. Everything in Base plus the advanced Learn module (certificates
- learning journeys
- training programmes
- curricula)
- Surveys
- Innovation Management
- Premium — enterprise plan
- up to 13.00 EUR per user and per month at the 250-user tier. Everything in Plus plus app integrations
- career management
- the Talent Dashboard
- the exit risk matrix
- the 9-box grid
- performance evaluation
- individual development plans
Data, GDPR & hosting
A consolidated view of how GFoundry handles your data.
GDPR overview
Implementation is documented in concrete terms. The data protection policy, version 1.3 updated in June 2026, is expressly built on Regulation (EU) 2016/679 and names GFoundry Lda as controller, with VAT number PT510809510. It lists the lawful bases relied on — consent, contract performance, legal obligation and legitimate interest — and the full set of data subject rights, including portability. Breaches affecting clients are notified within 72 hours. All data is processed and stored in the European Union; transfers outside it rely on the European Commission's Standard Contractual Clauses and on the Data Privacy Framework. Google Analytics runs with IP anonymisation. A Data Protection Officer is referenced in the GFoundry Intelligence policy. One gap remains: no customer-facing Data Processing Agreement is published or explicitly offered.
Who owns the data?
Two layers must be told apart. Under section 6 of the Terms of Service, all content, software, graphics and other materials provided by the Service belong to GFoundry Lda or its licensors, protected by copyright, trademark and intellectual property law. Customer data is a separate matter: GFoundry Lda acts as data controller for its own products and services, while the use of the platform itself is governed by a distinct Service Agreement that defines contractual obligations and the processing of client data. The AI layer never receives personal data — the vendor states that the model sees tokens, not names, that identity resolution happens server-side, that each tenant is isolated, and that customer data is never used to train models.
Reuse rights
The Terms of Service grant no reuse licence over the platform's own materials: content, software and graphics remain the property of GFoundry or its licensors, and users must limit themselves to lawful use that respects the rights of others. On their own data, customers keep practical control. GDPR rights include portability in a structured format, module dashboards export results to XLS, the Gi Talent agent exports its analyses to Excel, and a final backup can be requested before deletion at the end of the contract. Content produced by the Learning Designer agent is described as ready to publish inside the customer's own platform. Beyond that, no published clause states whether or how AI-generated material may be reused commercially outside GFoundry — the documents simply do not address it.
Data retention & training
Hosting summary
Hosting runs on Google Cloud Platform, described in the help centre as a secure multi-tenant SaaS infrastructure built on Kubernetes, Cloud SQL and Cloud Storage. The vendor states that all data is processed and stored in the European Union, and that client uploads such as PDFs and learning materials stay in EU-located cloud infrastructure; no individual country is named, only the region. Storage is encrypted and transmission secured over HTTPS and SSH keys, with redundant storage advertised at eleven nines of durability, continuous monitoring and automatic recovery. Each customer's data is isolated from the others, and the AI layer runs in per-customer isolated instances. Only essential, anonymised prompts are transmitted to the two named AI subprocessors, OpenAI OpCo LLC and Google's paid-tier Gemini, both covered by Standard Contractual Clauses for EEA and Swiss data. One caveat: the server answering for the public website is a Cloudflare anycast node geolocated in the United States, which reflects the marketing site's CDN and not where platform data lives.
Things to keep in mind
Risks and trade-offs to weigh before adopting GFoundry.
- The platform measures workplace behaviour continuously — pulse surveys, nine engagement metrics, exit-risk scores, a 9-box grid. Without a clearly explained purpose, that reads as surveillance and erodes the very trust it is meant to build.
- The Gi Talent agent sorts people into nine behavioural archetypes and raises overload alerts. Labels of that kind tend to stick, and a machine-assigned archetype can quietly bias promotion, staffing and development decisions.
- Leaderboards, badges and virtual coins can tip into unhealthy competition or gaming of the metric, especially when the scoreboard is built on commercial KPIs rather than learning.
- The vendor's own customer testimonial warns that gamification must be taken seriously and backed by a defined strategy; deployed without one, it produces novelty rather than engagement.
- Training content generated by AI from uploaded PDFs can carry factual errors or omissions. Nothing replaces a human review before a course is published to an entire workforce.
- Gi Practice scores interpersonal skills automatically after spoken role-play. Automated feedback on how someone handles a difficult conversation is a rehearsal aid, not a substitute for a manager's judgement.
- Exit planning deserves attention: enterprise plans commit for twelve months, client data is deleted after only a seven-day backup retention window at the end of the contract, and a single generic mailbox handles legal, GDPR and support requests alike.
Setup & Integrations
Technical difficulty
Two very different paths. The Base plan is self-service: an account is created from the website with no technical prerequisite announced. An enterprise rollout is a project. A first focused go-live takes a few weeks with a certified partner, whose consulting and integration hours are included in the Plus and Premium licences alongside initial training and an account manager; a setup fee applies. Single sign-on must be configured on the customer side. Integration effort scales with the route chosen: Make.com and Zapier need no developer, the REST API and webhooks need a technical team, and MCP needs a compatible AI assistant.
Deployment
Integrations
Behind GFoundry
Social
Resources
All the official URLs gathered for verification and reference.
Alternatives
Tools that compete with or complement GFoundry.
Frequently asked questions
What kind of organisation is GFoundry designed for?
Is there a free plan, and how much does the first paid tier cost?
Can I buy just one module?
What do the AI agents actually do?
Is my company's data used to train AI models?
Where is the data hosted?
How does GFoundry connect to the systems we already run?
How long does implementation take?
Are there mobile applications?
What happens to our data when the contract ends?
Should you pick GFoundry?
GFoundry is a mature product rather than a recent launch: the Portuguese vendor has been operating since 2014 and backs its claims with named case studies at DPD, Leroy Merlin, Carglass, Natixis, Randstad, Cork Supply and Pierre Fabre. Its distinctive move is one of positioning — it does not try to be the system of record. It layers gamified, AI-assisted experiences on top of whatever HR suite is already in place, which explains why several of its own comparison pages end by describing competitors as complementary rather than displaced.
The strongest argument in its favour is the data posture. Processing and storage sit in the European Union on Google Cloud, the AI subprocessors are named rather than hinted at, the substance of both Data Processing Agreements is published, and the vendor states plainly that personal data never reaches the model and that customer data is never used for training. For a European HR buyer, that level of disclosure is above the norm.
The reservations are practical. A 250-user minimum sold in blocks of 250, on a twelve-month cycle with no monthly payment, puts the complete platform out of reach for smaller organisations, and published prices are ceilings at the first tier rather than firm figures. There is no customer-facing DPA, and no public API documentation to evaluate the integration story before signing. Several published details also contradict each other — the registered address, the number of supported languages, the module count of the free plan — and the Terms of Service still date from 2023 while the privacy policy was refreshed in 2026.
The free Base plan, permanent for up to five users, is the sensible way to resolve that: it allows a real hands-on assessment before any commercial conversation begins.
- Choosing a selection results in a full page refresh.
- Opens in a new window.