Frequently Asked Questions

ContentCloud is a growing ecosystem of secure and compliant enterprise AI solutions for users in EU institutions and businesses. Today, the ecosystem includes our CCBot intelligent assistant for cited answers across content and data, and Memory for capturing private and governed knowledge work, with a broader direction toward reusable and governed knowledge operations.

Our platform can process structured and unstructured content, including PDFs, office documents, datasets and more, across multiple languages.

CCBot is a domain-grounded assistant that answers with citations from your sources (documents, datasets, websites). Unlike generic chatbots, it is constrained to your approved content and can be embedded into your portals and workflows.

Yes. We support multilingual corpora and can detect language automatically. You can configure supported languages for both the interface and content ingestion.

CCBot provides highly accurate responses because it only draws from your approved content sources. All answers include precise citations, allowing users to verify information. Response accuracy depends on content quality and completeness.

CCBot supports PDF, Word documents (.docx), PowerPoint (.pptx), Excel (.xlsx), plain text, HTML, and structured data formats. We can also ingest content from web pages and databases via API.

CCBot automatically detects when documents are updated and re-processes them to ensure responses reflect the latest information. You can configure update frequency and set up webhooks for real-time synchronization.

Yes. We design bespoke AI solutions that fit your information architecture and workflows, from pilots to integrations (e.g., Drupal modules, widgets and APIs).

We start with discovery (stakeholder interviews, content audit, risk and compliance review), then propose a pilot with measurable outcomes before scaling via integrations and enablement.

We offer EU cloud regions and private cloud or on-premise options. Data residency can be constrained to the EU, with encryption in transit and at rest.

We provide APIs, widgets and CMS connectors (e.g., Drupal modules) so you can embed search and Q&A, or push structured outputs (articles, FAQs, summaries) into your existing publishing flow.

We offer RESTful APIs for content ingestion, search queries, user management, and analytics. Our APIs support authentication via API keys or OAuth 2.0, with comprehensive documentation and SDKs for popular programming languages.

Yes. We support SAML 2.0, OpenID Connect (OIDC), and EU Login for seamless authentication. Enterprise customers can configure custom identity providers and access controls.

CCBot respects your existing permission structures. Users only see responses from content they have access to. We support role-based access control (RBAC) and can integrate with your directory services.

No. Your data is only used to serve your organization. We do not use your content to train shared models.

Documents are processed securely with access controls. Content can be stored in encrypted object storage or connected to your own storage. Retention and deletion policies are configurable.

Your bill has three parts: (1) a base monthly subscription for platform & hosting, (2) a variable ingestion fee when you add or update content, and (3) usage‑based inference metered by tokens (input/output). Annual discounts available; purchase orders accepted.

We offer Free, Usage‑based, and Enterprise/Custom plans. Start with Free for evaluation, scale with Usage‑based, or choose Enterprise for mission‑critical needs.

Yes. The Free plan supports up to 5 PDFs (≤20 MB each), 100 interactions/month, 1 project and 1 seat, session memory only, includes a “Powered by ContentCloud” badge, and community support. See Pricing for details.

Yes. Use the calculator on the Pricing page to simulate scenarios across subscription, ingestion, and token‑based inference.

Inference usage is metered by input and output tokens. Longer prompts, larger context windows or longer answers consume more tokens.

Subscriptions are billed monthly or annually. We provide quotes on request and accept POs for public sector and enterprise customers.

We support standard invoicing and bank transfers for enterprise/public sector. Card payments are available for self-serve tiers where applicable.

Book a demo with our team. We will review your requirements and propose a plan. Visit the Contact page to schedule a call.

We offer documentation, email support and dedicated account management for enterprise deployments.

Yes. We provide onboarding sessions, editor playbooks and optional workshops for communications and knowledge teams.

We follow privacy-by-design principles and align with GDPR. Options include EU data residency, encryption in transit and at rest, and private cloud or on-prem deployments.

We design our services around security management practices aligned with enterprise requirements, alongside GDPR obligations. The exact controls, hosting arrangements, and supporting documentation may vary by deployment model and are reviewed during solution design and procurement.

Yes. We recommend a scoped pilot on priority content with clear success criteria (accuracy, coverage, time-to-publish) before scaling to more use cases.

Most pilots run 4–8 weeks depending on content volume, languages, integration needs and compliance reviews.

Memory is a private memory tool for knowledge work. It captures selected activity from sources you choose — browser history, files, clipboard, email, chat, photos, and voice — across your devices, and puts it in one searchable, AI-powered timeline. Start with private recall, then extend into team memory and organizational continuity through governed sharing.

Memory Personal is your private recall — events captured from your chosen sources, stored locally, synced to your account. Memory Work adds workspace boundaries, admin-defined source policies, automatic sensitivity classification, and a promotion workflow for sharing curated knowledge with your team or organization.

No. Raw events remain user-scoped by default. Admins can define which sources and channels are allowed, set retention rules, and review quarantined high-sensitivity events — but they do not have default access to a user's raw timeline. The enterprise layer only receives artifacts that have been explicitly promoted through a review workflow.

Teams benefit from the knowledge that surfaces through promotion workflows — decision logs, handover notes, summaries, and knowledge cards. These curated artifacts carry the value of operational memory without exposing raw personal capture.

Proprietary LLM memory features store your data on the provider's servers in opaque formats with no meaningful export. Memory stores your data locally on your devices and on EU-hosted infrastructure you control. You can export your entire dataset at any time in standard formats, use any LLM provider for queries, and you are never locked in.

Memory works on macOS, Windows, Linux (via the Memory Sidekick desktop app), iOS, Android (via the Memory Mobile app), and via the Memory Portal web dashboard. All devices sync to a single unified timeline.

Browser history (Chrome, Chromium, Firefox), watched folders, clipboard, Mattermost chat, IMAP email, photos, voice memos, and personal notes. Each source is independently configurable per device.

Yes. Memory includes AI-powered Q&A that answers questions about your timeline in plain English with citations back to source events. Semantic search is available on Pro and Business plans. AI providers are configurable (Mistral, OpenAI, or self-hosted).

Yes. Memory is built on local-first architecture with EU-hosted sync (Hetzner). Sensitive content is automatically classified and quarantined. Users retain full export and deletion rights at all times. Your data is never used to train shared models.

Yes. Full export in JSON and CSV formats is available at any time. GDPR-compliant data export is also available, scoped by user or workspace if needed.

Yes. Memory includes Docker Compose deployment profiles and full self-hosting documentation. All workspace, enterprise, classification, and audit features are enabled by default on self-hosted installations. EU-hosted (Hetzner) managed deployment is also available.

Yes. Workspace admins can create source allowlist policies (which sources can push events), channel filter policies (which channels are captured), and per-workspace retention rules specifying how long events are kept and what action to take (delete, anonymize, or archive).

Memory is currently in pilot and available free of charge to early access participants. Pricing plans (Free, Pro, Business) will be announced when Memory reaches general availability. Pilot users will receive priority access to future plans.

Memory is currently available through an invite-only pilot program. You can request early access via the Contact page or the Memory product page. Pilot participants get full feature access at no cost and a direct feedback channel with the team.