SvaraCall — Acceptable Use Policy

Sarwagyna Private Limited Effective Date: 14 September 2026 Version 1.0

This Acceptable Use Policy ("AUP") forms part of the SvaraCall Terms of Service. Capitalised terms have the meaning given there. Breach of this AUP is a material breach of the Terms and may result in immediate suspension or termination without refund.

We enforce this AUP at our discretion. We may investigate suspected breaches, review call samples and campaign configurations, and report criminal conduct to law enforcement.

---

You must not use the Services to:

1.1 Deceive. Impersonate a government body, bank, regulator, court, law enforcement agency, hospital, or any person or organisation you are not. Misrepresent the identity of the calling party. Create false urgency about legal, financial, or medical consequences.

1.2 Defraud. Operate or support advance-fee fraud, loan or job scams, fake investment or trading schemes, pyramid or multi-level marketing recruitment, lottery or prize scams, phishing or vishing, or any scheme to obtain money, credentials, or personal data by deception.

1.3 Harass. Place repeated calls intended to annoy, intimidate, threaten, or abuse. Continue calling a person who has asked you to stop. Use abusive, threatening, or obscene language. Conduct calls designed to shame a debtor to their family, employer, or community.

1.4 Coerce or extract data unlawfully. Elicit passwords, OTPs, PINs, card numbers, CVVs, banking credentials, Aadhaar numbers, or other authentication credentials from an End User.

1.5 Target protected contexts. Place automated or AI voice calls to emergency service numbers, hospital emergency lines, patient rooms, ICUs, paging services, or any number where automated calling is prohibited.

1.6 Interfere. Attempt to disrupt, overload, probe, or circumvent the security of the platform, our Subprocessors, or any telecommunications network. Conduct unauthorised penetration testing. Bypass rate limits, quotas, or spend controls.

1.7 Misuse voices. Clone, synthesise, or use the voice of a real person without that person's documented and specific consent. Use a synthesised voice to impersonate a public figure, celebrity, or any identifiable individual.

1.8 Circumvent suppression. Use the Services to call numbers registered on the National Customer Preference Register or any applicable do-not-call, do-not-disturb, or internal suppression list, other than where a lawful exemption applies and you can evidence it.

1.9 Conduct prohibited activity. Promote or facilitate unlawful gambling or betting, unregulated financial products or deposit schemes, unlicensed lending, unapproved pharmaceuticals or medical claims, tobacco or narcotic substances, weapons, adult content, or any activity unlawful in the jurisdiction where the call is received.

1.10 Spread harm. Distribute malware, disseminate content that incites violence or hatred against any group, spread content prohibited under Indian law, or conduct disinformation campaigns.

1.11 Compete unfairly. Use the Services, their outputs, or synthesised audio to train, benchmark, or develop a competing voice AI product. Resell, redistribute, or expose the Services as a standalone product to third parties without a written reseller agreement with us. Reverse engineer, decompile, or attempt to derive the source or model architecture of the platform.

1.12 Scrape. Use automated means to extract data, voice samples, model outputs, or platform content beyond your own account data through documented APIs.

---

You must:

2.1 Identify the business on whose behalf the call is made, and the purpose of the call, at the start of every outbound call.

2.2 Configure your AI Agent to respond truthfully when an End User asks whether they are speaking to a machine, an AI, a bot, or a recording. Instructing an AI Agent to deny being AI, or to claim to be a named human employee, is a breach of this AUP.

2.3 Obtain and retain evidence of a lawful basis for contacting each End User, and for recording where recording is enabled.

2.4 Announce call recording at the start of the call where recording is enabled.

2.5 Scrub calling lists against applicable suppression registers before each campaign and at the intervals required by regulation.

2.6 Configure your AI Agent to recognise and act on opt-out language in the languages you operate in — including phrasings equivalent to "stop calling me", "remove my number", "do not call again" — and to end the call politely on request. Add such numbers to your suppression list within seven (7) days.

2.7 Provide a working path to a human being for End Users who ask for one, in any use case involving complaints, disputes, medical matters, or financial obligations.

2.8 Respect calling hours. Do not place outbound commercial calls before 9:00 a.m. or after 9:00 p.m. in the End User's local time, or within any narrower window required by regulation or by the End User's stated preference.

2.9 Register as a sender or telemarketer on the DLT platform of an access provider, and use registered headers and approved templates, where your communications require it.

2.10 Keep your knowledge base accurate. You are responsible for what your AI Agent tells people. Do not deploy agents that make claims about prices, eligibility, medical outcomes, returns, or legal rights that you cannot substantiate.

---

3.1 Healthcare. You may not use the Services to deliver a diagnosis, a treatment recommendation, a triage decision, or any clinical instruction. Appointment scheduling, reminders, follow-up logistics, and administrative queries are permitted. Do not configure agents to collect clinical history without a documented lawful basis and appropriate safeguards.

3.2 Financial services. You may not use the Services to solicit deposits, sell investment products, or provide investment advice unless you hold the relevant RBI, SEBI, or IRDAI registration and can produce it. You must comply with sectoral codes on recovery agent conduct, including restrictions on calling hours, frequency, and third-party contact.

3.3 Debt collection. Calls must be limited to the debtor. You must not disclose the existence of a debt to family, neighbours, employers, or references. You must not threaten legal action you are not entitled to take or do not intend to take. You must not misrepresent the consequences of non-payment.

3.4 Education. Where the End User is a minor, you must have verifiable consent from a parent or guardian and must comply with the DPDP Act's restrictions on processing children's data, including the prohibition on behavioural tracking and targeted advertising.

3.5 Insurance. Mis-selling, misrepresenting policy terms, and inducing lapse or replacement of an existing policy by misdescription are prohibited.

---

Do not submit to the Services, and do not configure your AI Agents to capture, any of the following without our prior written agreement:

• payment card numbers, CVVs, expiry dates;
• bank account numbers, IFSC combinations sufficient to transact, net-banking or UPI credentials;
• OTPs, PINs, passwords;
• Aadhaar numbers, PAN, passport, driving licence, or other government identifiers;
• biometric templates;
• health records, diagnoses, or prescription data;
• any category of data subject to heightened statutory security requirements.

If an AI Agent inadvertently captures such data, notify us at security@sarwagyna.com so we can assist with purging it.

---

We may impose concurrency limits, per-minute call rate limits, daily volume caps, and API rate limits, and may adjust them. We may throttle or suspend accounts whose traffic threatens platform stability, triggers carrier-level blocking, or generates abnormal complaint or spam-flag rates.

Sustained high abandonment rates, very short average call durations combined with high volume, or elevated spam-flagging of your numbers are treated as abuse signals and may trigger investigation.

---

Report suspected abuse of the Services to abuse@sarwagyna.com. Report security vulnerabilities to security@sarwagyna.com. We do not pursue legal action against good-faith security research that does not access other customers' data, does not degrade the Services, and is reported to us before disclosure.

---

Depending on severity we may: warn you; require remediation within a stated period; throttle or pause specific campaigns; suspend telephony capability; suspend the account; terminate the account; release your numbers; retain data as required for investigation; report to carriers, regulators, or law enforcement.

We are not obliged to monitor. Our failure to act on a breach is not a waiver of our right to act on it later.

Suspension for AUP breach does not entitle you to a refund.

---

Sarwagyna Private Limited | contact@sarwagyna.com | +91 6305036991