1.1. These Terms govern your use of Frosty Agent (the “Service”), an AI sales and support agent that you deploy on your own website and messaging channels. Frostrek LLP (“Frostrek”, “we”, “us”, “our”) provides the Service.

1.2. By creating an account or using the Service you agree to these Terms. If agreeing on behalf of a company, you confirm authority to bind it.

1.3. You must be at least 18 and capable of entering into a contract. The Service is for business use only.

1.4. These Terms incorporate our Acceptable Use Policy and Data Processing Addendum. Our Privacy Policy is a notice describing how we handle personal data.

1.5. Order of Precedence: (a) Data Processing Addendum; (b) signed enterprise order form; (c) these Terms; (d) Acceptable Use Policy.

2.1. Frosty Agent provides an AI agent that answers visitor queries, captures leads, books meetings, prepares quotations, and hands conversations to human staff on web widgets, WhatsApp, and connected channels.

2.2. License grant: We grant you a non-exclusive, non-transferable, worldwide right to access and use the Service for your business during an active subscription.

2.3. We provide the software and hosted infrastructure. You remain responsible for what your agent says, the accuracy of your uploaded knowledge base, and your customer relationships.

• 3.1. Credentials: You must keep account credentials secure and not share accounts across individuals.
• 3.2. Compromise notice: Notify [email protected] promptly if you suspect unauthorized access or credential compromise.
• 3.3. Team members: You are responsible for all actions taken by team members you invite to your account.

• 4.1. Your Content Ownership: Knowledge bases, catalogues, and branding assets remain your intellectual property. You grant us a limited license solely to host and run the Service for you.
• 4.2. Customer Data Roles: You are the Data Controller for your visitors’ conversations and leads; we act as your Data Processor.
• 4.3. Privacy Warranties: You warrant that you have lawful basis, necessary consents, and proper privacy notices to deploy the agent to your audience.

• 5.1. Model Nature: The agent generates responses using statistical large language models. It can make mistakes or hallucinate despite built-in guardrails.
• 5.2. Verification: You are responsible for configuring guardrails and reviewing your uploaded knowledge base.
• 5.3. Prohibited uses: Must not configure for medical, legal, financial advice, or automated credit/employment decisions without human review (Section 1 AUP).
• 5.4. Emergencies: The agent is not an emergency service. English crisis detection routes to public help lines, but cannot be relied upon as a clinical safeguard.
• 5.5. Quotations: AI-drafted quotes are indicative estimates until formally approved and executed by your staff.

6.1. We may update or route across upstream AI models and providers to ensure high availability, speed, and latency performance.

6.2. Our active list of infrastructure sub-processors is maintained in our Privacy Policy Section 12.

• 7.1. Third-party terms: Connecting external integrations (e.g. WhatsApp, Google Calendar) subjects you directly to those providers’ terms and policies.
• 7.2. WhatsApp & Meta Rules: For WhatsApp, you carry sole responsibility for obtaining verifiable opt-in consent, respecting messaging windows, and adhering to Meta’s task-specific assistant rules effective January 15, 2026.
• 7.3. Calendar integrations: You are responsible for authorization scopes and verifying appointment conflicts.

• 8.1. AUP incorporation: Use is subject to our Acceptable Use Policy.
• 8.2. Notice to cure: For suspected non-critical breaches, we provide 3 business days to cure. Immediate suspension applies only for platform security threats, illegal conduct, or regulatory mandates (with notice provided within 24 hours).
• 8.3. Targeted isolation: Where technically feasible, we suspend only the affected agent or integration channel rather than the whole account.
• 8.4. Appeal SLA: Appeals submitted to [email protected] are reviewed by an independent team member within 5 business days. If wrongly suspended, fees are credited.

9.3. Non-billable: Conversations where the agent produces no answer or system errors occur on our side are not charged. WhatsApp messages are charged only upon successful delivery.

9.4–9.9. Pricing & Overage: Prices are in INR ex-GST. Credits expire at end of billing cycle. Overage can be enabled with spend caps or disabled to auto-pause. Price changes require 30 days’ advance notice.

9.10–9.11. Razorpay Mandates & RBI Rules: Recurring renewals notify by email 7 days before monthly and 30 days before annual renewals. Banks issue 24h pre-debit notices under RBI guidelines. Turn off auto-renewal anytime in dashboard.

• 10.1. Cancellation: Cancel anytime in dashboard or via [email protected]. Access continues until billing cycle end.
• 10.3. 14-Day Money-Back Guarantee (Quarterly / 6-Month / Annual plans): If you cancel within 14 days of first payment having consumed no more than 20% of one month’s allowance, request a 100% refund. Excludes setup/add-on fees. (Monthly plans cancel anytime with no lock-in).
• 10.6. SLA: Refund decisions are rendered within 5 business days, and funds remitted to original payment method within 5 to 7 business days of approval.

• 11.1. Availability: We target 99.5% uptime during each calendar month, excluding scheduled maintenance.
• 11.2. Support Hours: Mon–Fri, 10:00 to 19:00 IST at [email protected].
• 11.3. Maintenance: Planned maintenance is communicated at least 48 hours in advance and scheduled during low-traffic windows.

12.1–12.3. Output Ownership: All rights in the Frosty Agent platform belong to Frostrek. You own the outputs, drafted quotes, and summaries generated by the agent for your account.

13.1–13.4. Both parties agree to maintain mutual reasonable care over non-public information. Confidentiality obligations survive termination for a period of 3 years (trade secrets survive indefinitely).

14.1–14.3. Either party may terminate on 30 days’ notice for uncured material breach.

14.4–14.5. 30-Day Export Window: On termination, live data is retained for 30 days for self-service export, then purged (backups overwritten within rolling 60 days). Statutory invoices are retained for 8 years.

15.1–15.3. We warrant providing the Service with reasonable skill and care. Except as expressly stated, the Service is provided “as is” without implied warranties of merchantability or fitness for a particular purpose.

• 16.1. No limits on fraud, death, or personal injury caused by negligence.
• 16.2. Exclusion of indirect, special, or consequential damages.
• 16.3–16.4. Liability Cap: Total aggregate liability of either party over any 12-month period is capped at the greater of fees paid in that period or ₹1,00,000.
• 16.5. Uncapped Items: Caps do not apply to payment obligations, indemnities under clause 17.1, confidentiality breaches (clause 13), IP infringement, AUP breaches, or DPA/data protection breaches.

17.1. Customer indemnity: You indemnify us for third-party claims arising from your content, AUP violations, unlawful instructions, or unreviewed AI output decisions.

17.3. Provider IP indemnity: We defend and indemnify you against third-party IP infringement claims directly alleging the core Frosty Agent software infringes copyright, trademark, or patent.

18.1–18.2. We provide 30 days’ advance notice by email for material changes. Prior versions are archived with changelog dates.

• 19.3. Assignment: 30 days’ advance notice with termination right for assigned party.
• 19.5. Force Majeure: Excusable delay up to 30 days with mutual termination and prepaid fee refunds thereafter.

• 20.1–20.2. Jurisdiction: Governed by laws of India; exclusive jurisdiction in courts at Gurugram, Haryana.
• 20.3. Consumer Protection: Consumer Protection Act 2019 rights preserved.
• 20.4. Commercial Mediation: Pre-institution mediation under Section 12A of Commercial Courts Act 2015 required before litigation.

Appointed officer under Consumer Protection (E-Commerce) Rules 2020 and DPDP Act 2023: