Twig AI

Twig AI is designed for rapid, low-friction deployment, often requiring no involvement from the customer's internal IT or engineering teams.

• Deployment Method: Cloud-based, fully managed from a user interface (UI).
• IT Requirement: Described as having "Fast deployment with zero IT support"; the customer experience (CX) team can often manage the setup from the UI.
• Implementation Time: Generally quick, with marketing statements suggesting the platform can launch in 30 days for new clients.
• Onboarding Support: Available through structured service tiers (e.g., Managed Onboarding) that provide specialized assistance with initial setup and configuration.

The platform offers extensive controls that allow users to fine-tune the AI's behavior, tone, and the data sources it relies on to match specific business processes and brand persona.

• Persona and Tone: Users can define the AI's specific persona, including the desired answering style, tone, and response length (e.g., short, long, bulleted).
• Data Source Control: Provides fine-grained control allowing agents to manually select, un-select, or add specific documentation sources when generating an answer for a customer.
• Action Integration: Supports API calls to enable the AI to convert customer requests into actions, such as fetching user-specific information (e.g., order status) or initiating workflows.
• Chatbot Branding: The embedded AI chatbot can be customized to align with the look and feel of the company's website or application.

Pricing is typically custom for enterprise customers and often structured around a tiered subscription model combined with a potential usage-based component for AI resolutions.

• Pricing Structure: The official pricing is custom (requires a demo) and tailored to the company's specific use case and team size.
• Subscription Models: Tiers can range from basic self-serve to comprehensive managed outcome plans.
• Usage Costs: Some reported plans indicate a cap on included AI Resolutions per month, with resolutions exceeding that limit incurring an additional per-resolution fee (e.g., $0.50 - $0.99), leading to variable monthly billing.
• Managed Service Fees: Costs for support (maintenance/support) are often bundled into the higher subscription tiers, which include services like dedicated AI Specialist setup and weekly quality check-ins.

Twig AI offers a blend of dedicated expert support, platform-based analytics, and a design that minimizes the need for internal agent training.

• Implementation Support: Managed service tiers include direct human assistance from an AI Specialist for setup and ongoing quality management.
• Agent Productivity Tools: The platform features a Quality Workbench and detailed analytics to help admin users monitor usage, capture feedback, and refine AI performance.
• Technical Support: Availability of 24/7 Live Support Online is often offered for technical issues under various service level agreements (SLAs).
• Agent Training Impact: The system acts as a co-pilot within existing ticketing systems (like Zendesk or Outlook), requiring minimal to no new training for support agents.

Security architecture is robust, utilizing advanced techniques like PII anonymization and RAG to ensure data privacy, integrity, and compliance with enterprise security standards.

• PII Anonymization: Twig sanitizes ingested customer data (e.g., support tickets, documentation) by applying Personally Identifiable Information (PII) filters to remove sensitive details before the data is used to train or generate responses.
• Data Integrity: Utilizes a Retrieval-Augmented Generation (RAG) architecture to provide deterministic answers that are cited to trustworthy source documents, preventing AI hallucinations.
• Authentication: Supports Single Sign-On (SSO) protocols, including Google SSO and enterprise-grade SSO, for robust user access control.
• Penetration Testing: The platform is reported to have passed multiple Fortune 100 penetration tests.

The system is self-improving, with automated data synchronization and continuous learning loops that ensure the AI's knowledge base and response quality remain up-to-date.

• Data Refresh Rate: Data pipelines are configured to automatically refresh from connected private data sources (Confluence, docs, etc.) on a scheduled basis, often weekly or based on customer configuration.
• Continuous Learning: User and agent feedback on AI-generated responses are captured and trigger workflows to refine and improve the AI model's accuracy.
• Synthetic Data Generation: The platform generates new synthetic data and QnA from support ticket history, which continuously enriches the knowledge base to handle new or complex queries.

Twig AI's policies are aligned with major data protection laws, implying that the customer retains ownership of their proprietary business data and maintains standard rights over personal data.

• Customer Data Ownership: The customer typically retains ownership of their proprietary documentation, support ticket history, and other internal business data ingested by Twig.
• Privacy Compliance: The processing of personal data is handled in accordance with standards like the General Data Protection Regulation (GDPR).
• Portability Rights: Users maintain the right to data portability, allowing them to request a copy of their personal data in a structured, commonly used, and machine-readable format.
• Data Control: The system's PII removal process ensures that customer data used for model generation is privacy-safe, giving the customer control over sensitive information.

Scaling flexibility is primarily driven by subscription tier upgrades, which unlock higher limits on the volume of AI-resolved interactions.

• Scaling Up: To handle increasing support volumes, customers move to higher-tiered plans that include a significantly greater number of AI Resolutions per month.
• Immediate Scaling: In the event of a sudden spike in demand, the usage-based component of the pricing model allows for automatic scale-up by incurring the predefined per-resolution overage fee.
• Scaling Down: Scaling down is achieved by moving to a lower subscription tier, typically during contract renewal, to reduce the monthly base cost and align the plan with consistently lower support volumes.

Specific contractual obligations for renewal and cancellation are client-specific and found in the Master Service Agreement (MSA), though general terms regarding termination for cause exist.

• Contract Terms: Specific renewal periods (e.g., annual, multi-year) and cancellation notice requirements are defined in the private client contract (MSA) and are not publicly listed.
• Termination for Cause: The service terms allow Twig to suspend or immediately terminate the user's license and access to the services if the user is found to be in violation of the core agreement or terms of service.
• Service Changes: Twig reserves the right to update the services or discontinue features, with the customer's contract likely outlining notification protocols for such changes.

Twig AI maintains essential compliance certifications and follows stringent data processing regulations required to operate in enterprise environments.

• Security Compliance: The platform asserts that it is Soc 2 compliant (or operating under strict controls toward this certification), a critical auditing standard for security, availability, processing integrity, confidentiality, and privacy.
• Data Regulation Compliance: Adherence to international data privacy laws, notably the General Data Protection Regulation (GDPR), is mandatory for their European operations and widely applied to data processing.
• Internal Governance: Compliance is supported by the platform's PII filtering and its commitment to providing traceable, cited answers (RAG architecture), which aids in meeting audit requirements for data accuracy and security.