You are interacting with AI
The Indigloo Assistant, supported chatbot demonstrations, AI writing tools, and Agent Studio responses are generated by software, not a human adviser. An agent name, persona, voice, or job title describes a configured role; it does not identify a real employee, licensed professional, or independent decision-maker.
Current hosted text demonstrations use NVIDIA. A technology described on a capability or product page may instead refer to a proposed client implementation. Provider names, model versions, integrations, availability, and limits can change. A separate client agreement defines the implementation actually delivered.
Accuracy, sources, and professional decisions
AI can invent facts and citations, misunderstand instructions, omit important details, reproduce bias, or produce outdated and unsafe suggestions. A confident answer or a displayed source is not proof that a statement is correct. Check the original source, permissions, and current context.
Do not rely solely on these tools for medical, legal, financial, employment, eligibility, compliance, or security decisions. They are not a replacement for qualified advice. An authorized person must review consequential decisions and any generated code, communications, recommendations, or reports before use.
Chatbots and knowledge-grounded responses
The public assistant helps explain Indigloo services. Its answers are not binding quotations, promises, service-level commitments, or authorization to act for Indigloo. Confirm commercial information through info@indigloo.co or an accepted client agreement.
Retrieval-augmented generation (RAG) and document grounding are implementation capabilities, not a promise that every public answer is connected to verified private documents, real-time records, or current web results. A chatbot may provide a plausible answer without having access to the source you expect.
AI agents, demonstrations, and actions
Agentic AI describes systems that can be designed to plan or use tools. Public capability descriptions do not mean a hosted demo has every illustrated integration or can act autonomously. Only explicitly implemented, enabled, and authorized functions are available.
AI Coworker is a text-planning and drafting demonstration. Approving its plan produces a draft; it does not authorize or cause independent browsing, filesystem access, purchases, message sending, or external workflow execution.
Where a separately delivered agent can take real actions, its operator must define least-privilege access, authorized scope, approvals for sensitive or irreversible steps, monitoring, and a way to stop or recover. A model's suggestion is not user permission. These website declarations do not grant any external-system access.
Prompts, uploaded content, and retention
Only submit information you have authority to share. Prompts, relevant conversation context, and supported voice inputs can be processed by the disclosed providers. Do not enter live credentials, personal information, client secrets, confidential documents, payment details, or regulated data into public demonstrations.
Do not assume input is automatically anonymized or that all AI use is private, uncached, unlogged, or excluded from provider retention or training. Application histories, activity records, and caching vary by feature; provider terms and any applicable client processing agreement also matter. See the Website Privacy Policy before submitting content.
The Security Analyzer does not persist submitted analysis content or reports in its own application flow or semantic cache. This does not promise that the provider retains nothing. Its password sample checker is browser-local; never enter a real password.
Voice and speech features
Voice features require the permissions and controls shown by the feature and browser. Browser or operating-system speech services may process speech remotely under their own terms. A hosted speech turn can transmit audio and transcripts to NVIDIA, with browser speech used where supported as a fallback.
Synthetic speech is not a human speaking live. Recognition may mishear words, accents, names, or numbers; check transcripts and outputs. Do not record other people without the required notice and permission, and do not send sensitive audio to a demonstration.
Security, code, OCR, and estimates
Security tools provide text-based AI guidance, not live URL reputation checks, infrastructure scans, breach detection, penetration testing, or proof that a system is secure. Test only systems and materials you are authorized to assess. Independently validate findings and remediation.
Where document extraction, OCR, translation, summarization, or estimation is offered in an enabled feature or client implementation, omissions, transcription errors, and incorrect assumptions are possible. Check results against originals. ROI estimates and examples are illustrative, not guarantees of savings or results. A listed feature may require separate setup or may not be enabled in the public demo.
Intellectual property, models, and third-party licenses
Company-owned application code, branding, and original materials remain proprietary. Third-party models, libraries, assets, APIs, and hosted services retain their own rights, licenses, acceptable-use rules, and attribution requirements. A feature notice does not replace a required license or grant distribution rights.
AI-generated output may be non-unique, resemble third-party material, or be subject to rights or restrictions. We do not guarantee originality, non-infringement, copyright eligibility, exclusivity, or a complete chain of rights. Review generated text, code, images, and other content before publishing, selling, deploying, or redistributing it.
Dependency inventories describe reviewed evidence and scope; they are not blanket commercial-use or redistribution clearance. Describing a model or library does not imply ownership, endorsement, or permission to redistribute it. Client ownership and delivery rights must be addressed in the applicable agreement.
Availability, misuse, and contacting a person
Demonstrations are subject to access rules, quotas, provider outages, and changes. Do not use them for impersonation, fraud, credential theft, unlawful surveillance, unauthorized testing, or attempts to defeat safeguards. Do not treat a successful demo as a production-readiness, compliance, or uptime certification.
For a human response, an incorrect output, a privacy concern, an accessibility barrier, or a rights complaint, contact info@indigloo.co. Do not include passwords or sensitive evidence in an initial report. Separate signed agreements govern professional services; these declarations do not replace professional legal review.
Questions about this notice?
Contact us at info@indigloo.co.