Back to list
Annexes

ANNEX 1 – Service Description (SaaS)

Version v2.1Last updated June 9, 2026

ANNEX 1 – Service Description (SaaS)

Service

Lyyli.ai is a communication workflow management system for organizations' internal and external communication. Lyyli combines brand management, content production, project management, and AI-assisted workflows in a single platform. Lyyli is not a text generator; it manages the full communication process from ideation to approval and publication.

Key Features

AI Chat and Planning

• Workspace- and project-level AI chat conversations

• Full message history

• Chat forking and sharing

• Archive and bulk chat operations

• Dedicated planning mode: separate planning area with planning-specific conversations

• AI-assisted communication planning with direct output to content management

Content Management

• Content library organized by workspace and project

• Full content CRUD (create, read, update, delete)

• Version history and content restore

• Bulk updates

• Comments and attachments on content items

• Review workflow with public review links

• External commenting on shared content without a user account

Image Management

• Image library with tagging

• AI image generation

• Prompt enhancement for images

Meeting Intelligence

• Meeting recording and diarized transcripts

• Per-meeting insights (summary, key topics, action items, sentiment, recurring themes)

• Cross-meeting theme analysis

• AI-generated action proposals

• Pending actions view

• Meeting Agent settings and configuration

• Calendar integrations: Google Calendar and Outlook Calendar (requires environment configuration)

Inbox

• Unified inbox with unread counts

• Reply threads

• Create content directly from inbox briefs

• Inbound email processing

Projects and Workspaces

• Multi-workspace support

• Project-level organization within workspaces

• Workspace board views with configurable columns (Kanban)

Web Ideas and Tracking

• Workspace-level web source tracking

• Automated content idea refresh via scheduled processing

Admin and Settings

• Organization admin dashboard

• User profile settings

• Home page preference

• Locale selection

• Role-based access: Admin, Editor, Viewer

Features in Development (not yet available as live user features)

The following features are implemented in the codebase but not yet exposed in the navigation or available to all users:

Carousel Studio: visual carousel content creation

Brand Kit: centralized brand asset management

These will be communicated separately upon release.

Integrations and Third Parties

The service integrates with the following third-party platforms:

Microsoft Teams and Slack: bidirectional integration for idea collection and content distribution

LinkedIn: content publishing with Customer authorization

Google Meet, Microsoft Teams, and Zoom: Meeting Agent participation

Google Calendar and Outlook Calendar: Meeting Agent scheduling (requires environment configuration)

Email (inbound): inbox processing for ideas and briefs

Third-party services are not under Lyyli.ai's control. Lyyli.ai is not responsible for the availability or changes of these services, including API interface changes. When a user connects an integration, Lyyli.ai processes data only with the user's authorization and solely to provide the service.

Data Security and Privacy

Zero Data Retention (ZDR): Lyyli uses exclusively AI model providers operating under Zero Data Retention policies. Customer content is not stored by AI providers and is not used for model training. Data is processed in memory in real time and destroyed immediately after the API response is returned.

EU-only infrastructure: Application data is stored in the EU. All data routing occurs within European borders in compliance with GDPR.

Encryption: Data encrypted in transit (TLS 1.3) and at rest (AES-256).

Request anonymization: AI requests are anonymized; individual users cannot be linked to specific AI requests.

Organizational isolation: workspaces are isolated at the organizational level.

Audit trail: all content actions are logged and traceable.

DPA: Data Processing Agreement is a standard built-in feature.

• No customer data is used to train Lyyli's own or third-party language models.

Full details: lyyli.ai/trust

Limitations and Terms of Use

Content Responsibility: The Customer is responsible for the accuracy and approval of content, and for ensuring that content published through Lyyli.ai to third-party platforms complies with those platforms' rules and terms of use.

Authorizations: Automation and integrations operate only with authorizations provided by the Customer.

Meeting Agent: The Meeting Agent joins meetings via a meeting URL and notifies participants via chat that AI-assisted note-taking and transcription is active. The Customer is responsible for ensuring that meeting participants are appropriately informed in accordance with applicable law and organizational policy.

External review links: Public review links allow external parties to comment on content without a Lyyli account. The Customer is responsible for managing access to shared review links.

Integration Suspension: Lyyli.ai reserves the right to block or suspend the use of an integration if its use endangers data security, violates third-party terms, or causes harm to the service.

Reasonable Use: Reasonable use limits are defined on a plan-specific basis. See current limits at lyyli.ai/pricing/compare. The service may not be used for automated mass messaging (spam) or other activities that violate platform terms.

Onboarding and Support

Starter: onboarding available as a separate one-time fee (199 €)

Growth: onboarding included

Professional: onboarding included; Named Account Manager and priority support with 99.5% SLA

• Support contact: hello@lyyli.ai

• Documentation and trust information: lyyli.ai/trust and lyyli.ai/legal

Updates

Continuous development; no separate version updates to the Customer's environment. New features and AI models are made available to all users as they are released. Change management in accordance with Annex 4 principles.