Questions and answers about DailyOne

DailyOne is web-based school software that can also be installed as a PWA. It brings school organisation, teaching, communication, support and administration together so schools need fewer isolated tools and information remains connected to its educational context.

DailyOne is designed for schools, school authorities, municipalities and education organisations. Dedicated roles and views support school administrators, teachers, students, parents or guardians, and platform administration.

A single-purpose app usually covers a timetable, chat or files. DailyOne connects those areas with class and subject management, attendance, substitutions, digital classrooms, parent meetings, learning tasks, translations, AI drafts, support services and secure role controls.

Schools can centrally manage users, teachers, students, guardians, classes, subjects and rooms. Timetables, assessments, special notices, digital lessons, cancellations, attendance and substitution suggestions use the same structured school data.

Teachers find timetables, subject chat, files, tasks, attendance and student assignments in one place. They can organise learning material, translate documents and prepare editable AI-assisted drafts for quizzes, exercises, assignments and assessments before reviewing and publishing them.

Yes. Recurring processes such as student import, account credentials, class and subject assignment, timetable planning, sick notes, attendance, substitutions, file organisation, parent appointments and school messaging become connected digital workflows.

DailyOne includes role-based user management, classes and subjects, timetables, rooms, attendance, sick notes, substitution support, direct, class, subject, parent and staff chats, files, learning materials, tasks, quizzes and learning games, digital exercises, assessment drafts, elections, parent meetings, video and audio services, translation and local AI.

Students see their timetable, digital lessons, materials, tasks, quizzes, messages and attendance information in a clear interface. Bilingual content, simplified explanations and additional language support help learners follow lessons when language barriers exist.

DailyOne includes a protected help area with anonymised one-to-one audio connections to suitable contacts or other students. It can create a low-threshold route to conversation, but it does not replace medical, psychological or emergency crisis support.

Parents or guardians receive their own accounts and a parent chat. Schools can plan parent meetings online or in person, allocate time slots, manage invitation status and start protected video calls directly from digital appointments.

Teachers or schools define the date, start time, duration, format and participating families. Appointments are organised per child and guardian, and online conversations can be linked directly to an assigned video room.

DailyOne supports system and additional languages, bilingual text display and translations of learning materials and school content. The source language can remain visible while the required translation is added directly below it.

New and multilingual students gain easier access to timetables, tasks, communication and lesson materials. Language support is built into everyday school processes instead of being separated into additional, difficult-to-access tools.

Digital education is inclusive only when information and functions are accessible regardless of disability, language or device. DailyOne considers semantic structure, screen readers, keyboard access, visible focus, high-contrast themes, colour-vision modes, large navigation and right-to-left languages.

DailyOne costs EUR 2.20 before VAT, or EUR 2.62 including 19% German VAT, per student and billed month. Schools are billed for 12 months per year. The current calculator on the Pricing page transparently shows monthly and annual totals for the selected number of students.

Yes. The sustainable minimum is EUR 120.00 before VAT per billed month and applies below 55 students. From 55 students, the regular calculation is the student count multiplied by EUR 2.20 before VAT.

Only student accounts are billed. Teacher, school administration and management accounts are included in the published platform price, so schools do not pay a separate licence fee for every additional staff role.

12 monthly payments or one annual payment. Both options cover all 12 months.

For 100 students, the regular price is EUR 220.00 before VAT per billed month and EUR 2,640.00 before VAT for 12 months. For 500 students, it is EUR 1,100.00 per billed month and EUR 13,200.00 before VAT for 12 months.

For operation on an organisation’s own infrastructure, the pricing calculator provides a separate setup calculation. Hardware, networking, backups and day-to-day operation generally remain the organisation’s responsibility; optional update and support services are shown separately.

DailyOne Local AI uses a language model that can run locally through llama.cpp. Supported content therefore does not need to be sent to public AI clouds. AI can assist with document analysis, translation, quizzes, tasks, assessment drafts and administrative drafts while results remain editable and subject to human review.

Local AI can summarise and translate existing documents, prepare editable tasks, quizzes, exercises and assessment drafts, and create structured proposals for administrative workflows. Teachers and authorised staff can reduce preparation work while retaining professional control and final approval.

No. DailyOne’s core functions and intended local AI do not require a mandatory connection to OpenAI, Claude or another public AI cloud. The local runtime can operate inside the controlled environment of the school or responsible operator.

No. AI produces reviewable drafts and does not independently decide grades, promotion, support needs, disciplinary action or educational opportunities. Responsible people must review, edit and explicitly confirm relevant results.

DailyOne is not designed to use school inputs or outputs to train general public AI models. Local processing and controlled data paths are intended to prevent supported content from being sent unnecessarily to public AI services.

DailyOne is designed around data-minimising and controlled school processes, including roles and permissions, school and tenant separation, server-side checks, protected access, local AI processing and legal documentation. Actual GDPR compliance also depends on deployment, contracts, hosting, configuration and use by the responsible school or authority.

The platform is designed for European and self-determined operation. Schools can choose a managed service or their own server environment. Local AI, controllable data paths and no mandatory public AI cloud reduce external dependencies.

Communication is tied to authenticated roles, approved contacts, schools, classes or subjects. API checks, password hashing, protected sessions and a WebCrypto-compatible chat foundation complement organisational separation. No serious platform can promise absolute security, so DailyOne uses several layers of protection.

A draft data processing agreement and recipient register are available under Legal. The register names Hetzner for German hosting and IONOS as an email configuration reference. This is not yet a complete supplier list confirmed for every school: legal entities, services, locations, access and further subprocessors must be recorded and approved for each installation. Registration does not conclude the agreement.

The operational schedule describes hosting generally in the payer’s country according to the operator, generally with Hetzner for Germany. The actual data centre and locations for production data, backups and support access must be specified in the operational schedule and recipient register. A German supplier address or a country selected during onboarding does not establish the actual storage location.

Yes. The legal package includes a draft DPIA and data flow, security and risk documents. The responsible organisation must complete and assess them for its modules, data, people, recipients and safeguards. The scope, timing and cost of personal assistance must be requested separately; a completed assessment for every school is not promised.

The processing agreement and operational schedule provide for data return or export and subsequent deletion under an agreed procedure. Data scope, original files, structured formats, deadlines, secure recipients and completeness checks must be agreed; backups and remaining retention obligations are recorded separately. The reviewed documents do not establish special insolvency protection or guaranteed continued operation. This must be clarified expressly before contracting.

No. The documents require a specific retention schedule by data type and operating model. User deletion, school-year reset and end-of-contract deletion are separate processes. Backups, downloads, external recipients and records that must be retained need their own rules and evidence. Immediate complete deletion of every copy is not promised.

The documents describe managed SaaS and self-hosting. Server AI runs in the configured server environment; optional device AI runs on the individual device. The operational schedule assigns infrastructure, updates, backups, keys, support and AI operation to the supplier or appointed operator. Local does not automatically mean on school premises or on a teacher’s device.

The documented Qwen3 4B Q4_K_M model file is 2,497,280,256 bytes, about 2.5 GB. This is only the file size, not a minimum RAM specification. Memory and computing requirements also depend on runtime, context length, concurrent use, translation services and model choice; the alternative Ollama path can use qwen3:14b. The reviewed documents do not specify a universally validated minimum configuration. Capacity must be tested with representative tasks and expected load.

The operational schedule has fields for support channels, staffed hours, priorities, response and recovery targets, escalation, maintenance windows and notice periods. These must be agreed individually. The template expressly creates no blanket 24/7 or 99.9 percent commitment. A response time is also different from the time needed to resolve an incident completely.

DailyOne documents published changes in its release history; the app displays the ten latest releases under News, Fixes, Features. The reviewed documents do not establish a binding dated public roadmap or a fixed schedule for major changes. Required features, change notifications, training and maintenance planning should be agreed separately for deployment.

No. The operational schedule requires a specific backup plan for databases, files, keys and configuration, plus documented recovery tests. A backup script or hosting option does not prove that all records are actually backed up and recoverable. Backup intervals, retention, acceptable data loss and recovery targets must be agreed and verified.

Qwen3 4B Q4_K_M is documented for llama.cpp and the intended device AI. There is also a configurable Ollama path with the default name qwen3:14b and optional approved external providers. Dynamic translation additionally uses LibreTranslate and the local MADLAD fallback google/madlad400-3b-mt. The operator must establish which path and model version actually run in the specific installation; DailyOne AI does not mean the same model everywhere.

A selectable language or complete interface dictionary does not establish the quality of free-text translations. The reviewed documents include technical translation tests but no general expert quality approval for every required language. Have competent speakers assess your language pairs using typical parent letters, subject texts and tasks. Names, numbers, negations and long texts deserve particular attention; important communications need the original and a way to ask questions.

The payer’s AI administration manages external API and MCP connections and their school assignments. No such connections may remain approved for a school intended to use local AI only; the backend checks access against approvals. This concerns DailyOne integrations. It does not completely block external AI websites, other apps or existing school network access. Additional device and network rules from the operator are required for that.

The payer’s authorised AI administration configures connections and school approvals. According to the operating documentation, API keys remain in the instance vault; teachers and school administrators cannot retrieve or export them. Technical approval does not replace assessment of purpose, recipient, contract, processing location and data scope.

DailyOne makes no such blanket guarantee for external services. Storage, training, disclosure and deletion depend on the actual provider, agreement and settings. These points must be clarified before approval. Statements about the local DailyOne inference path must not be applied to external API or MCP services.

No. In device mode the model request runs locally, but source data may first be loaded from the DailyOne server. Saved drafts, accepted results and AI activity events may still be exchanged with DailyOne. Local model processing and the offline app shell do not promise that every school function works without a network.

DailyOne documents local demo and test accounts; the draft contract accommodates demo and test access using synthetic data. The reviewed documents do not establish a generally bookable pilot with a fixed duration or published pilot price. Use Contact to request access, duration, scope, assistance and costs. Trial access alone creates neither a paid production contract without agreement nor a general promise of a free pilot.

The reviewed project documents do not establish an approved public list of production reference schools with contacts. Demo schools and test accounts are not customer references. Ask DailyOne for an approved reference and an introduction with that school’s consent; availability of a conversation cannot be promised here.

The reviewed documents do not establish blanket state approval or independent certification of DailyOne. The processing agreement, technical and organisational measures, DPIA template, model card and security concept are reviewable documents, not official authorisations or certificates. Request any claimed evidence with issuer, scope, product version and validity. Approval of the particular school deployment must be clarified separately.

DailyOne includes help and support content; the legal package includes an AI literacy and training concept and human oversight guidance. The concept covers safe inputs, privacy, error detection and checking AI outputs. Specific sessions, formats, included services and prices are not generally fixed and must be requested. A training concept is not evidence that training has taken place.

The operational schedule covers contracting parties and authority, the processing agreement and schedules, recipients and locations, retention, roles, approved modules and verification of the actual installation. Tests use synthetic data. Digital onboarding creates organisational records and access on the existing installation; it does not automatically provision a new server at the selected location.

Local server AI processes requests in the server environment configured for DailyOne. For self-hosted installations, the respective operator is responsible for that environment. Supported options include a local llama.cpp runtime with Qwen3 4B Q4_K_M and a configurable Ollama runtime. The default Ollama model name is qwen3:14b. The actual model, version and processing location must be documented for the installation. These local inference routes do not require a public AI cloud service. DailyOne does not provide for training general-purpose models with school inputs or outputs in these routes. Model and software downloads are separate from teaching requests. Hosting, infrastructure access, logging and backups depend on operating arrangements. Local does not automatically mean processing on the device or within a particular country.

In device mode, the application sends requests to the installed local runtime on the same device. The intended runtime uses Qwen3 4B Q4_K_M. This does not make the school platform an offline application: source data may previously have been retrieved from DailyOne, and saved drafts, adopted outputs and content-free AI activity events may still be transmitted to DailyOne. The school and device operators govern access protection, updates, disk encryption, shared devices and downloaded files. Browser data, exports and operating system copies are not erased merely by ending a request or deleting a school account. Check the selected operating mode before a confidential request.

The payer administers its external providers and their technical school assignments. The school checks purpose, data scope, lawful basis and suitability. Selecting an external API service and initiating an applicable AI function causes DailyOne to transmit the assembled request to the approved endpoint. This may include inputs, selected document or teaching content and technical connection data. MCP connects tools and is not itself an AI model. Approved tools may retrieve data or change an external system. The documented MCP workflow requires preview and explicit confirmation. Check recipients, data and effects before initiating it. A later cancellation cannot reliably reverse a transmission or external change that has already occurred. Before approval, establish providers, roles, contracts, subprocessors, locations, retention, deletion and possible provider training purposes in writing. Third-country transfers also require Chapter V GDPR compliance, including necessary safeguards and supplementary measures. Technical availability or a provider setting does not establish compliance. Personal data must not be transmitted without a valid basis. DailyOne cannot itself guarantee an external provider’s actual storage, training, onward disclosure or deletion. The payer must agree and review appropriate contracts and account settings. The school and provider retain their own statutory and contractual duties.

These notices describe AI assistance for teaching materials, questions, translation, document preparation and administrative drafts requiring approval. Availability depends on role, school assignment, installation and authorisation. Enabling a function does not establish compliance with data protection or education law. The SaaS contract, data processing agreement, privacy notices and school rules also apply. The school or school authority responsible under applicable education law determines school purposes and means. DailyOne processes school data on instructions within the agreed scope where it acts as a processor. Separate provider purposes, such as contract administration or necessary operational security, must be identified separately. Funding the service does not automatically make a payer the controller of all school data.

The school documents an appropriate basis under Article 6 GDPR together with applicable education and state law. Special categories of personal data also require a condition under Article 9 GDPR. Acceptance of usage rules or acknowledgement of privacy information is not blanket consent to AI, international transfers or subsequent processing. Where consent is exceptionally used, it must be freely given, specific, informed, demonstrable and withdrawable, with a reasonable alternative without disadvantage. Age, capacity, guardianship and Article 8 GDPR where applicable require assessment for the particular use. Mandatory school activities cannot use forced consent to replace a missing lawful basis.

Transmit only necessary information. Remove names, identifiers and other identifying attributes where the task does not require them. Pseudonymised information remains personal data where identification is possible. Documents may also contain information about uninvolved people. Health information, counselling content, particularly confidential student records, passwords and access keys must not routinely be entered into general AI fields. Exceptions require documented lawful grounds, approval and safeguards. Check copyright, usage rights and confidentiality obligations for third-party materials.

The external provider supplies its service under the payer’s contract. DailyOne and Deep Nodes Ai IT Solutions UG (limited liability) do not warrant that provider’s availability, accuracy or independent data processing. Their own statutory and contractual duties, including secure integration, authorisation and data processing, remain in place. Mandatory liability, including under Article 82 GDPR and for intent, gross negligence or injury to life, body or health, is not excluded.

AI outputs may contain false facts, unsuitable tasks, biased representations or incorrect translations. Teachers review suggestions for subject accuracy, language and suitability for learners. Consequential translations require the original and an opportunity to seek clarification. Change drafts requiring approval and MCP calls follow their documented preview and confirmation process; the backend checks permissions and school assignment. Some DailyOne quizzes and exercises automatically calculate feedback and scores against stored solutions. Answers, results and progress may be stored. This differs from an AI model independently deciding school grades. Incorrect questions or solution keys can also distort rule-based feedback and must be open to correction. Grades, promotion, support needs, educational access and similarly significant decisions must not be adopted from AI suggestions or practice scores without substantive review. A responsible person must independently assess relevant circumstances, be able to change results and consider objections. Merely formal confirmation is insufficient. Article 22 GDPR and education law require separate consideration. The school and provider document purposes, models, recipients, approvals and risks, support appropriate AI literacy and review material changes. Where processing is likely to result in a high privacy risk, the controller conducts a DPIA under Article 35 GDPR before processing; high residual risk requires prior consultation under Article 36.

Classification depends on intended purpose and actual use. AI evaluating learning outcomes, steering learning based on those outcomes or deciding educational access may fall under Annex III of the AI Act. Local operation and human involvement do not by themselves establish an exception. An Article 6(3) exception requires documented assessment of the case; profiling may exclude it. According to the European Commission’s published information, AI literacy duties and existing prohibitions have applied since 2 February 2025, and Article 50 transparency duties since 2 August 2026. The AI Omnibus, which entered into force on 27 July 2026, postpones the relevant high-risk rules for Annex III until 2 December 2027 and regulated products under Annex I until 2 August 2028. This does not postpone the GDPR or prohibitions already in force. Under Article 111(4), systems placed on the market before 2 August 2026 have a limited transition until 2 December 2026 for Article 50(2). Eligibility must be evidenced for each system; other transparency duties are not generally suspended. Before a new or materially changed use, the provider and school document their respective roles, risk classification, purposes, oversight, information duties and evidence. Depending on the case, conformity requirements, registration, monitoring and a fundamental rights impact assessment may apply. These notices do not replace assessment of the particular use or certify general compliance.

DailyOne is not authorised for social scoring, biometric emotion recognition in schools, covert behaviour monitoring or manipulative exploitation of children’s vulnerabilities. These purposes must not be introduced through prompts, external tools or later changes of purpose. A prohibition does not guarantee technical prevention of every form of misuse. Schools establish age-appropriate rules, supervision, accessible contacts and a process for distressing or discriminatory outputs. AI replaces neither psychological counselling nor emergency assistance. Staff working with AI need sufficient competence on limitations, privacy, error detection and human oversight.

Model outputs, imported documents and external tool responses may contain misleading instructions. A check of format, permissions or origin does not establish accuracy. Approved HTTPS endpoints and separate credential administration must be complemented by secure configuration, updates and operation. The general server AI job service temporarily holds requests and results in memory. Unattended jobs are scheduled for cancellation. Finished results are removed from this job store by the next successful cleanup after five minutes. This does not apply indiscriminately to every business function. Cancellation or releasing a memory reference guarantees neither physical memory overwriting nor erasure of data already sent to an external service. Adopted outputs, teaching materials, translations and confirmed changes may create persistent business records or files. School retention and erasure rules apply. Downloads, device storage and copies held by authorised recipients must also be considered.

For school processing, individuals may contact their school or its data protection officer and request access, rectification, erasure, restriction, portability or object where statutory conditions apply. Consent can be withdrawn for the future. The right to complain to a competent data protection authority remains unaffected. Contacts and responsibilities are given in the respective privacy notices. Incorrect or adverse results can be challenged with the responsible teacher or school. Privacy and security incidents must be reported without undue delay through agreed channels so that suspension, preservation of evidence, remediation and necessary notifications under Articles 33 and 34 GDPR can be assessed. The provider assists as required by law and contract. Changes to purpose, model, recipient, data scope or operating mode require renewed assessment. Applicable school rules, use decisions and contracts must be retained in a traceable form. An account confirmation without a separate document version link does not establish the exact content of every document displayed at the time. Relevant law includes the GDPR, Regulation (EU) 2016/679, and the AI Act, Regulation (EU) 2024/1689 as amended. The Commission’s overview of application dates is available at https://digital-strategy.ec.europa.eu/en/policies/regulatory-framework-ai. The educational basis depends on the state law applicable to the school.

The Log button under AI and Downloads opens the account’s own AI history. School administrators can also view events for their school. Records contain the account identifier, time, function, mode, source, correlation identifier and status. The display may resolve the current account name. This history does not store prompts, model responses, document content or credentials. Device events are labelled as reported by the device; DailyOne cannot independently attest to local execution. AI history supports accountability, troubleshooting and secure operation. It is not authorised for unrelated performance monitoring. The school documents its lawful basis for using logs. For separate processing for operational security, the provider must determine its own basis, including a balancing assessment where required. AI activity events are visible for 90 days. The scheduled cleanup runs at server startup and then hourly, removing older events when it completes successfully. Failures may delay actual erasure; the operator must monitor and remedy them. Account deletion removes related activity events from the production database. Protected backups follow separate backup and erasure arrangements. The 90-day period applies only to this AI activity history. Teaching outputs, materials, business change records and separate administrative records for external AI connections may have different requirements. Their periods and erasure processes must be determined by purpose. Automatic deletion of all records after 90 days is not promised.