Description of software services - Gyde (Flip Learning)
Last updated: 06.08.2026
gyde provides the following functions for the creation of learning content, for delivering e-learning, microlearning and seminars, and for knowledge management on the web and in native apps:
Learning content and authoring tools
Courses and interactive learning content. Learning content is structured as course → module → chapter → lesson → slide and supports more than 17 interactive element types, including formatted text, image, video, audio, file, links, single- and multiple-choice quizzes, drag-and-drop sorting, cloze text, free-text and numeric inputs, sliders, calculation and scoring elements, and cross-references.
Course creation (content management system, “CMS”). Authors create, edit and arrange the complete course hierarchy. Courses are published as versioned snapshots; editing locks prevent concurrent changes, and individual modules or entire courses can be copied and pasted.
SCORM import. SCORM 1.2, SCORM 2004 and cmi5-compliant course packages can be imported and played via an integrated, standards-based player. The import function is provided via a sub-processor established in the United States (Annex 3 to the DPA); the transfer is governed by Sections 3.3 and 10 of the DPA. The Customer may have the function deactivated at tenant level; where it is not activated, no processing takes place via it. Courses created in gyde use a native format and do not rely on SCORM. Irrespective of the import function, courses created in gyde can be exported per language version as SCORM 2004 (4th Edition) or cmi5; the export function can be enabled per tenant and is enabled upon the Customer’s request.
AI-assisted functions
AI assistant. A dialogue-based assistant carries out natural-language instructions across the entire platform, not only creating and editing course content but also day-to-day administration such as managing users, teams, seminars and registrations. It acts via the platform’s own interface exclusively within the requesting user’s permissions, is aware of the object currently open and responds in the user’s language. EU-hosted models are generally used, prompts are processed in the EU, and no customer data is used to train the models (exceptions are possible for the AI video studio, which the Customer can enable optionally). The AI functions can be disabled, their use is logged exclusively for troubleshooting purposes, and the provider-side requirements of the EU AI Act are met.
AI-assisted course creation from documents. A complete course, including structure, slides and elements, can be generated from a source document (e.g. a PDF), a knowledge page or a topic brief. Each suggestion is displayed as a preview and is only created after express approval by the author. Uploaded source documents are stored exclusively in the customer environment and are accessible only to the Customer; temporary files are deleted after processing. The documents are not used to train AI models. Operation and infrastructure are certified to ISO 27001.
AI image generation. Images for courses, modules and slides can be generated and added via a preview-and-approval flow.
Automatic translation. Course content can be translated automatically into the configured target languages; more than 30 content languages are supported.
Knowledge management
Knowledge base. Administrators maintain hierarchical knowledge spaces and pages with rich content, which are published to learners as versioned snapshots, with semantic and full-text search across all content.
AI knowledge assistant. A dialogue-based assistant answers questions exclusively on the basis of the organisation’s published knowledge base, with mandatory source references; general model knowledge is not used. The answers are reconciled with the stored facts by grounding; optionally, the assistant can be configured to output only source references instead of generated answers.
Assessments and certification
Exams and assessments. Structured assessments with information and question slides validate learners’ knowledge and support multiple attempts.
Certificates. Certificates defined by administrators are issued automatically when configurable criteria (course progress and attendance) are met, or awarded manually. Each certificate contains a tamper-proof hash and a generated PDF and can be verified publicly without logging in; external issuing bodies are supported.
Seminars and in-person training
Seminar and in-person management. gyde manages trainer-led training end to end, as standalone seminars or linked to a course. Administrators create, publish, reschedule, cancel and duplicate seminars, and individual dates can be rescheduled or cancelled independently. Registered learners and trainers can export a seminar’s schedule as a standards-compliant .ics calendar file.
Seminar series and automatic scheduling. Recurring programmes are organised into seminar series with shared metadata, pricing defaults and an iCalendar schedule. New seminars are generated automatically in advance from a series template, with configurable lead time and optional automatic publication.
Delivery formats. Each date is online, in person, hybrid or asynchronous, with a dial-in link for online dates and room and location details for in-person dates. Rooms are managed as bookable resources with capacity, equipment, address and time zone.
Trainer management. Trainers maintain their own availability and block calendars; the system automatically detects room and trainer scheduling conflicts when a date is saved or rescheduled and can assign trainers based on the required qualification and minimum competence level. Trainers have self-service access to their schedule, participant lists and attendance recording.
Registration and enrolment. Seminars support open, approval-based, invitation-only or closed registration with internal or external (public) visibility. Participants register themselves, are added by an administrator or imported in bulk via CSV; an approval-and-rejection workflow governs approval-based seminars, and participants can unenrol themselves up to a configurable deadline. External registrants can be invited via time-limited, revocable links and register via a public endpoint; with each registration, a price and VAT snapshot is recorded for documentation purposes.
Waiting lists and group bookings. If a seminar is fully booked, registrations are placed on a waiting list and moved up automatically as places become available. Group bookings reserve a contingent of places with a purchase-order (PO) reference and a nomination deadline for individual participants; possible scheduling conflicts for a participant can be checked before the registration is confirmed.
Custom registration forms. Each seminar series can contain a configurable registration form with text, email, consent, single-choice, multiple-choice and scale fields. Forms follow a draft-and-publish lifecycle, each published version is stored as an immutable snapshot, and the submitted responses are recorded at registration.
Attendance and completion. Attendance is recorded per date and participant in six statuses (present, absent, excused, partial, no-show, late cancellation), with optional notes, recorded minutes and a check-in time, and can be entered individually or in bulk. Completion can be based on course progress, an attendance threshold or manual confirmation, and language-dependent attendance lists are generated as PDF.
Public seminar catalogue. Customers can publish catalogue pages that present seminar series and standalone seminars as cards with available places, price, schedule, location and ratings, each with its own registration form.
Seminar materials and process checklists. Files can be attached to a date, a date template or a series, with separate visibility for participants or trainers, and seminar-related task checklists track preparatory and follow-up steps with owners and status.
Video and microlearning
Microlearning (Flashes). Short, mobile-optimised videos are distributed to teams as microlearning units, with completion tracking.
AI video studio. Short learning videos can be generated from a text brief, as individual clips or as multi-part compositions combining AI video, text and image cards, AI-generated background music and designed subtitles. This uses Google’s globally operated video models (without EU data residency); the function is therefore deactivated by default and can be enabled optionally by the Customer. Once an EU-hosted variant is available, it will be offered.
Video-to-tutorial. A long training recording can be converted into a structured, chaptered and subtitled tutorial, with a manual review step before publication; this processing takes place in the EU.
Competence management
Qualifications and competences. Competences with proficiency levels, and roles with required competences and certificates, can be defined, with dashboards on competence gaps and role readiness.
Learning experience and end devices
Learning experience (web and mobile). The same interactive content is displayed identically in the web application and in the native apps, with a personalised home page, favourites, post-completion ratings, per-lesson notes and content-language switching.
Offline capability. Courses can be downloaded for offline use on mobile devices; media is cached locally, and learning progress is synchronised automatically on reconnection.
Organisation switching. Learners who belong to more than one organisation can switch between organisations in the app, with the respective branding applied at runtime.
Analytics and communication
Analytics console. Administrators, managers and trainers receive role-specific, metric-based dashboards on usage, completion rates, learning time, ratings and compliance. A minimum aggregation threshold protects against the de-anonymisation of small groups.
Exports. Participant lists, attendance and registration data can be exported as CSV; attendance lists per date as PDF.
Notifications. Users are informed via an in-app inbox, email and mobile push notifications, with per-user channel settings and customer-wide defaults.
Administration and integration
User and team management. Administrators create and import users (also in bulk), organise them into hierarchical teams with manager scope, and manage invite codes and registration-code documents.
Assignments. Courses are assigned to teams or individuals with mandatory flags and deadlines; content and learners can be tagged.
Integration of intranet, HR and process tools. Various intranets, HR and process tools can be integrated on request, via API (Application Programming Interface) or web view. In principle, any system that provides an API can be connected.
REST API and API clients. A documented REST API exposes the platform’s resources for integration and automation. Administrators can create and manage per-customer API clients, each authenticating with a purpose-bound, expiring API key; a Model Context Protocol (MCP) endpoint additionally makes the API available to authorised AI agents.
Custom agent skills. Administrators can define, publish and scope their own agent skills, reusable AI workflows that extend the assistant to automate recurring creation and administration tasks for selected audiences.
SLA and usage reporting. A report on the availability and use of the software service can be created, e.g. to check compliance with the SLA (Service Level Agreement).
Security, data protection and hosting
AI data residency and secure-AI mode. AI-assisted creation, translation, the knowledge assistant and video-to-tutorial processing run on EU endpoints; a customer-specific secure-AI mode restricts a customer to AI operated exclusively in the EU by disabling the AI video and music functions that are operated outside the EU.
Encryption and access control. Data is encrypted in transit; sensitive fields are additionally encrypted at column level, and access is controlled via fine-grained role-based permissions.
Hosting in the cloud. The service is operated in GDPR-compliant data centres in Germany (Google Cloud, Frankfurt region).
Retention and export
The retention periods for the individual data categories are set out in the table below. Retention periods marked as configurable can be set by the Customer within the limits provided for. Customer content can be exported in the formats and via the routes stated. In all other respects, deletion after the end of the contract is governed by the main agreement and the data processing agreement.
Data category | Retention | Configurable | Export |
Courses and knowledge pages | For the term of the contract | No | Courses: self-service export per language version as SCORM 2004 (4th Edition) or cmi5; the export function can be enabled per tenant and is enabled upon the Customer’s request. Knowledge pages: no self-service export; made available upon request via support |
Uploaded source documents | For the term of the contract | No | No self-service export; on request |
Temporary AI processing files | Deleted after processing | No | No export |
AI-generated media (image, video, audio) | For the term of the contract | No | No separate export; included in the course |
Automatic translations | For the term of the contract | No | No separate export; stored as course content in the target language |
Learning progress and results | For the term of the contract | No | Export via REST API (completion data) |
Certificates and verification hashes | For the term of the contract; public verifiability ends upon deletion | No | Automatic delivery as PDF to learners; download by administrators and managers; CSV overview of issued certificates in the administration area; public verification link; REST API |
Seminar, registration and attendance data | For the term of the contract | No | CSV (participation, attendance and registration data); PDF attendance list per date; schedule as .ics; REST API |
Registration form responses | For the term of the contract | No; published form versions are immutable | CSV (registration data) |
Data of external registrants | For the term of the contract; invitation links are time-limited and revocable | No | CSV (registration data) |
Trainer availability, block calendars and qualifications | For the term of the contract | No | Own schedule as .ics |
Price and VAT snapshots | The statutory retention periods of the invoicing party apply | No | No export |
AI assistant conversations | For the term of the contract | No | No self-service export; on request |
User profiles, attributes and team memberships | For the term of the contract | No | Export via REST API |
Usage statistics | Backend analytics for the term of the contract; anonymised reach measurement without a time limit | No | No export |
Search index and derived data | For the term of the contract | No | No export |
Deleted content (recycle bin) | For the term of the contract | No | No export |
Offline caches on end devices | No provider-side period; retention depends on the end device | No | No export |
Exports generated by users (CSV, PDF) | Generated on request and not retained permanently | No | CSV and PDF on request |
Provider system logs (no customer logs) | 1 year | No | No export |
Backups | Up to 12 months after deletion in the productive system | No | No export |
For the data categories marked above as “No self-service export”, the Provider makes the customer content available upon the Customer’s request via support in a common electronic format. The entries in the “Export” column refer to the export functions available during the term of the contract; the Customer’s rights under the main agreement (switching support) and under the data processing agreement remain unaffected. Following the end of the contract, the data listed above is deleted after expiry of the retrieval period, at the latest six (6) months after the end of the contract.
For the data categories marked as “No export”, no export function is available; where they contain Customer Content within the meaning of clause 1 of the Terms, the Provider makes it available upon request via support in a structured, commonly used and machine-readable format. Clause 9.5 of the Terms and the Customer’s rights under the Terms (export and switching support) remain unaffected.
System requirements
The following minimum requirements apply to the operating systems and browsers of end devices:
Operating systems
Windows: last 3 major versions
macOS: last 3 major versions
iOS: last 3 major versions
Android: last 6 major versions
Browsers
Chrome desktop & mobile: last 3 major versions
Edge desktop: last 3 major versions
Firefox desktop: last 3 major versions
Safari desktop: last 3 minor versions
Legal framework
This software service description sets out the characteristics of the gyde (Flip Learning) software service. Which functions are owed in a specific case follows from the offer in conjunction with the main agreement. gyde (Flip Learning) is provided by the Provider with the involvement of Gyde GmbH as a sub-processor; the processing of personal data is governed by the data processing agreement (DPA), in whose list of sub-processors (Annex 3) Gyde GmbH is included. For the optional AI video studio, processing may take place outside the EU; this function is deactivated by default and is only activated upon the Customer’s approval. The service quality and availability owed are governed by the Service Level Agreement for the Flip platform. The Provider may develop the service further within the scope of the contractual provisions. The current version of this software service description is available at https://www.getflip.com/legal/softwaredescription-gyde/.