relay privacy policy

effective date: september 28, 2026

thank you for using relay (“relay”, “we”, “our”, “us”, or the “assistant”), a personal ai call-screening service operated by rumik intelligence private limited (“rumik”).

this privacy policy explains how relay collects, uses, processes, stores, shares, and protects personal information when you use the relay mobile application, relay call-screening services, and related services.

relay is designed to answer and screen certain incoming calls on your behalf, understand who is calling and why, communicate with callers based on your preferences, and help you decide whether a call needs your attention.

we process personal information in accordance with applicable laws in india, including the digital personal data protection act, 2023 (“dpdp act”), the digital personal data protection rules, 2025 as applicable from time to time, the information technology act, 2000, and other applicable laws.

for the purposes of applicable data protection law, rumik intelligence private limited acts as the data fiduciary for personal data processed through relay where applicable.

1. about relay

relay is a personal ai assistant designed to handle and screen phone calls on your behalf.

depending on your preferences, device, carrier, and relay configuration, relay may:

  • receive calls forwarded to relay;

  • answer eligible incoming calls on your behalf;

  • identify itself as an ai assistant acting on your behalf;

  • speak with callers;

  • ask callers who they are and why they are calling;

  • transcribe what callers say;

  • determine whether a call may require your attention;

  • provide live updates while a call is being screened;

  • allow you to send instructions for relay to communicate to a caller;

  • allow you to take over a call;

  • generate a transcript after the call;

  • generate a call summary or digest;

  • maintain call history;

  • identify whether a caller appears in your uploaded contacts; and

  • provide notifications relating to calls and relay.

relay acts based on your settings, preferences, instructions, and the information available during a call.

ai-generated responses, summaries, caller classifications, urgency indicators, identity descriptions, and other outputs may occasionally be incomplete or inaccurate.

unless relay specifically states otherwise, information generated by relay should not be treated as independent verification of a caller’s identity, intentions, statements, or authority.

2. information we collect

2.1 account information

when you create or use a relay account, we may collect information including:

  • your name;

  • phone number;

  • authentication information;

  • account identifiers;

  • selected language;

  • selected relay voice;

  • call-handling preferences;

  • notification preferences;

  • subscription information;

  • device information; and

  • other settings associated with your relay account.

we use this information to:

  • create and maintain your account;

  • authenticate you;

  • personalise relay;

  • configure the assistant;

  • provide call-screening functionality;

  • communicate service information;

  • provide customer support;

  • maintain security; and

  • operate relay.

2.2 contact information

if you give relay permission to access your contacts, relay uploads relevant contact information to its systems.

this may include:

  • contact names;

  • phone numbers; and

  • information required to determine whether an incoming caller is already known to you.

we use uploaded contacts primarily to:

  • distinguish known contacts from unknown callers;

  • provide useful caller context;

  • apply call-screening preferences; and

  • improve the relevance of information shown during a call.

we do not use your uploaded contacts to independently advertise to those people or send them unsolicited messages.

if you revoke contact permission, relay will stop future contact synchronisation.

revoking device permission does not automatically delete contacts that were already uploaded.

previously uploaded contact information may remain associated with your account until:

  • you delete the relevant data;

  • you delete your relay account;

  • the information is no longer required to provide the service; or

  • another retention rule described in this policy applies.

after a valid account-deletion request, uploaded contact information will generally be deleted or irreversibly anonymised within 30 days, except where longer retention is required by law.

2.3 call information

when relay handles a call, we may collect and process information including:

  • caller phone number;

  • recipient phone number where required for routing;

  • caller name where available;

  • whether a caller matches an uploaded contact;

  • call date and time;

  • call duration;

  • call status;

  • call-routing information;

  • call-forwarding information;

  • whether relay screened the call;

  • actions performed by relay;

  • actions performed by you;

  • information provided by the caller;

  • instructions you send during live screening;

  • call outcome;

  • technical information relating to the call; and

  • information generated from the conversation.

we use this information to:

  • route calls;

  • operate call screening;

  • provide live updates;

  • generate transcripts;

  • generate summaries;

  • maintain call history;

  • troubleshoot problems;

  • prevent abuse;

  • provide customer support;

  • evaluate relay’s performance; and

  • improve relay as described in this policy.

3. call recordings

relay records and processes audio from calls handled by the assistant.

call recording data may include:

  • caller speech;

  • relay-generated speech;

  • audio transmitted during the relay-handled portion of a call;

  • recordings generated by telephony infrastructure;

  • transcripts generated from audio; and

  • information derived from the conversation.

call audio may be received, processed, or stored by rumik and by service providers used to operate relay, including plivo, livekit, and soniox, depending on the processing stage.

relay may use call recordings to:

  • understand the caller;

  • transcribe conversations;

  • generate responses;

  • determine why someone is calling;

  • create summaries;

  • identify whether a call may require your attention;

  • troubleshoot failed interactions;

  • investigate technical issues;

  • provide support;

  • evaluate assistant quality;

  • improve relay; and

  • improve rumik’s ai and voice systems in the circumstances described below.

relay is intended to record the portion of a call handled through relay’s infrastructure.

direct calls that do not pass through relay are not recorded by relay.

when you take over a call and relay no longer participates in the conversation, relay’s active ai handling of that call ends.

4. information about callers who do not use relay

people who call you may not themselves be relay users.

when relay screens their call, relay may process information about them including:

  • their phone number;

  • their name where available;

  • their voice;

  • statements made during the conversation;

  • information they voluntarily provide;

  • information inferred from the conversation; and

  • information generated by relay from that interaction.

relay processes this information to provide call-screening services to the relay user, maintain service security, create transcripts and summaries, and perform the other activities described in this privacy policy.

relay is designed to identify itself as an ai assistant acting on behalf of another person.

where appropriate or required, relay may also notify callers that the interaction is being recorded, transcribed, or processed using ai.

5. how relay processes a live call

relay uses multiple systems to perform different stages of a screened call.

not every provider receives the same information.

5.1 telephony and call routing — plivo

relay uses plivo for telephone numbers, pstn connectivity, call routing, forwarding, call control, and recording infrastructure.

plivo may process:

  • caller phone numbers;

  • recipient phone numbers;

  • call-routing information;

  • call timestamps;

  • call duration;

  • live call audio;

  • call recordings; and

  • technical call metadata.

more information about plivo is available at:

5.2 realtime communications — livekit

relay uses livekit as part of its realtime voice and agent infrastructure.

livekit may process information necessary to establish and operate a realtime voice session, including:

  • realtime audio;

  • session identifiers;

  • participant information;

  • connection information; and

  • technical session metadata.

more information about livekit is available at:

5.3 realtime speech-to-text — soniox

relay uses soniox for realtime speech recognition.

during a screened call, raw live caller audio may be transmitted to soniox so that spoken language can be converted into text.

relay may configure soniox with language settings appropriate to the selected relay language, including support for:

  • english;

  • hindi; and

  • hinglish interactions.

the resulting transcript is used by relay’s reasoning system to understand the conversation and determine what should happen next.

5.4 realtime reasoning — satin

relay uses satin as part of the realtime conversational reasoning pipeline.

satin may receive:

  • live transcript text;

  • previous conversational turns;

  • relay’s system instructions;

  • relevant call context;

  • relevant user preferences;

  • tool inputs;

  • tool results; and

  • other information necessary to determine relay’s next response or action.

satin does not need the original raw caller audio for this reasoning stage.

the caller’s speech is first converted into text by the speech-recognition system.

5.5 realtime speech generation — rumik silk

relay uses rumik silk to convert relay’s generated responses into speech.

silk includes rumik voice models such as:

  • mulberry; and

  • muga.

relay may also configure voices such as:

  • aisha;

  • ira;

  • lucas; and

  • theo.

silk primarily receives:

  • the text relay intends to speak;

  • selected voice information;

  • language information; and

  • speech-generation configuration.

muga and mulberry are voice-generation models and are not relay’s primary realtime reasoning systems.

rumik’s voice and agent infrastructure is also available through:

6. post-call transcription

after a call, relay may generate a durable transcript from the stored recording.

6.1 default transcription — self-hosted whisper

relay’s default post-call transcription system uses a self-hosted whisper implementation running within relay-controlled infrastructure.

when this configuration is used:

  • relay processes the stored call recording internally;

  • the recording does not need to be sent to an external transcription provider for this post-call transcription step; and

  • the resulting transcript can then be used to create the call summary and other call information.

6.2 optional transcription — soniox async

relay may alternatively be configured to use soniox for asynchronous post-call transcription.

when this configuration is enabled:

  • the stored call recording may be transmitted to soniox;

  • soniox processes the recording and returns transcript text; and

  • relay uses the transcript for downstream post-call processing.

the particular transcription system used may change as relay’s infrastructure develops.

7. call summaries and digests

after a call, relay may generate a summary or digest containing information such as:

  • who called;

  • why they called;

  • what they need;

  • whether the call appears to require your attention;

  • relevant details mentioned during the conversation;

  • requested follow-up actions; and

  • other useful information derived from the call.

relay may generate summaries using an openai-compatible llm integration.

depending on relay’s active infrastructure configuration, the summarisation provider may include:

  • openai;

  • google gemini;

  • groq; or

  • a self-hosted model operated by rumik.

only the provider configured for the relevant deployment or processing path receives the summarisation request.

if no external llm is configured, relay may use deterministic or internally hosted processing to generate certain post-call information.

7.1 what the summarisation provider receives

the summarisation system receives transcript text rather than the original call recording.

where technically feasible and not required for the summarisation task, relay may exclude separate structured account or telephony fields such as:

  • relay account identifiers;

  • caller-number metadata;

  • recipient-number metadata;

  • internal account information; and

  • separate recipient identity fields.

however, removing structured identifiers does not mean that a transcript contains no personal information.

a caller or relay user may say personal or identifying information during the conversation itself.

for example, the transcript may naturally contain:

  • a person’s name;

  • phone number;

  • address;

  • email address;

  • company name;

  • order number;

  • account reference;

  • appointment information;

  • delivery information;

  • financial information;

  • health-related information; or

  • other information voluntarily spoken during the call.

if that information is spoken as part of the conversation, it may appear inside the transcript sent for summarisation.

we therefore treat transcripts as potentially containing personal and sensitive information even when separate structured identifiers have been removed from the summarisation payload.

8. artificial intelligence processing

relay uses artificial intelligence and automated systems to:

  • transcribe speech;

  • understand conversational context;

  • determine how relay should respond;

  • generate responses;

  • generate speech;

  • understand caller intent;

  • classify calls;

  • identify whether a call may require your attention;

  • create summaries;

  • evaluate system quality;

  • detect failures; and

  • improve relay.

different systems perform different parts of this process.

for example:

  • plivo connects and routes the call;

  • livekit helps operate the realtime voice session;

  • soniox converts caller audio into realtime text;

  • satin processes conversational text and context to determine responses;

  • rumik silk, including muga and mulberry, converts relay’s response into speech;

  • self-hosted whisper or soniox may create the durable post-call transcript; and

  • a configured llm may process transcript text to generate the final call digest.

relay seeks to provide each system only with the information reasonably required for its specific function.

9. use of data for ai improvement

limited portions of information processed through relay may be used in certain circumstances to evaluate, train, fine-tune, test, or otherwise improve relay, rumik’s voice technology, and rumik-operated ai systems.

depending on the purpose, this may include limited portions of:

  • call audio;

  • transcripts;

  • generated responses;

  • call summaries;

  • user instructions;

  • model outputs;

  • corrections;

  • interaction outcomes;

  • failed interactions;

  • low-quality interactions; and

  • relevant technical metadata.

we may use this information to improve:

  • speech recognition;

  • english understanding;

  • hindi understanding;

  • hinglish understanding;

  • pronunciation;

  • voice quality;

  • caller-intent recognition;

  • conversational behaviour;

  • turn-taking;

  • model accuracy;

  • response quality;

  • safety;

  • latency;

  • reliability; and

  • call-screening performance.

authorised rumik personnel or contractors may review limited interaction samples where human evaluation is reasonably necessary for:

  • debugging;

  • quality evaluation;

  • annotation;

  • safety evaluation;

  • model evaluation;

  • model training; or

  • model improvement.

where reasonably possible, information that is not necessary for the relevant improvement task is removed, minimised, aggregated, masked, or de-identified.

9.1 third-party model training

sending information to a provider so that the provider can perform a service for relay does not by itself mean that the provider is permitted to use relay data to train its own models.

we do not sell call recordings, transcripts, uploaded contacts, or other personal information for ai training.

we do not permit unrelated third parties to use identifiable relay call data to train their own general-purpose or foundation models unless:

  • that use is separately disclosed;

  • it is permitted by our agreement with the provider;

  • it is permitted under applicable law; and

  • any consent required by applicable law has been obtained.

10. subprocessors and service providers

relay uses third-party and internally operated infrastructure to provide the service.

the main systems that may process call-related information include:

provider / system

purpose

information that may be processed

rumik intelligence private limited

rumik intelligence private limited

operation of relay and internal infrastructure

purpose: operation of relay and internal infrastructure

account information, recordings, transcripts, summaries, contacts, preferences, call data, and information required to operate relay

information processed: account information, recordings, transcripts, summaries, contacts, preferences, call data, and information required to operate relay

plivo

plivo

telephone numbers, call routing, forwarding, telephony, recording

purpose: telephone numbers, call routing, forwarding, telephony, recording

phone numbers, call metadata, live audio, recordings, routing information

information processed: phone numbers, call metadata, live audio, recordings, routing information

livekit

livekit

realtime communications infrastructure

purpose: realtime communications infrastructure

realtime audio, session information, participant information, connection metadata

information processed: realtime audio, session information, participant information, connection metadata

soniox

soniox

realtime speech-to-text and optional post-call transcription

purpose: realtime speech-to-text and optional post-call transcription

live caller audio and, where enabled, stored recordings

information processed: live caller audio and, where enabled, stored recordings

satin

satin

realtime conversational reasoning

purpose: realtime conversational reasoning

transcript text, conversational turns, relevant context, system instructions, tool inputs and results

information processed: transcript text, conversational turns, relevant context, system instructions, tool inputs and results

rumik silk / muga / mulberry

rumik silk / muga / mulberry

speech generation

purpose: speech generation

relay-generated response text, language and voice configuration

information processed: relay-generated response text, language and voice configuration

self-hosted whisper

self-hosted whisper

default post-call transcription

purpose: default post-call transcription

stored recording processed within relay infrastructure

information processed: stored recording processed within relay infrastructure

configured summary llm

configured summary llm

generation of post-call summaries

purpose: generation of post-call summaries

transcript text, which may contain personal information spoken during the call

information processed: transcript text, which may contain personal information spoken during the call

firebase

firebase

authentication, notifications, application infrastructure and diagnostics where configured

purpose: authentication, notifications, application infrastructure and diagnostics where configured

authentication data, device information, notification tokens, identifiers and technical metadata

information processed: authentication data, device information, notification tokens, identifiers and technical metadata

mixpanel

mixpanel

product analytics

purpose: product analytics

product events, identifiers, device information and usage metadata

information processed: product events, identifiers, device information and usage metadata

payment providers

payment providers

subscriptions and payments where applicable

purpose: subscriptions and payments where applicable

transaction information, payment status, subscription information and related billing metadata

information processed: transaction information, payment status, subscription information and related billing metadata

not every provider receives information from every relay interaction.

the providers involved depend on:

  • which relay feature is being used;

  • the active infrastructure configuration;

  • whether optional processing systems are enabled; and

  • the information necessary to perform the relevant function.

11. product analytics

relay may use analytics services such as mixpanel to understand how people use the application and how well features perform.

analytics may include information such as:

  • onboarding completion;

  • screens viewed;

  • call-screening started;

  • call-screening completed;

  • call-screening failed;

  • actions taken during live screening;

  • notifications opened;

  • settings changed;

  • feature usage;

  • app errors;

  • crashes;

  • performance information; and

  • application version.

raw call audio and private call transcripts are not intended to be included in ordinary product-analytics events.

12. firebase and notifications

relay may use google firebase for functionality including:

  • authentication;

  • push notifications;

  • device registration;

  • diagnostics;

  • application infrastructure; and

  • related services.

depending on the functionality used, firebase may process:

  • account identifiers;

  • authentication information;

  • device information;

  • notification tokens;

  • technical metadata; and

  • diagnostic information.

13. payments and subscriptions

where relay offers paid subscriptions, payments may be processed through third-party payment providers or application-store billing systems.

relay does not need to store your complete debit-card, credit-card, upi, or bank-account credentials itself.

we may receive information including:

  • transaction identifiers;

  • payment status;

  • subscription plan;

  • purchase date;

  • subscription start date;

  • renewal information;

  • expiry date; and

  • limited billing metadata.

payment providers process payment information according to their own terms and privacy policies.

14. permissions relay may request

contacts

relay requests contact access to identify whether an incoming caller is someone already known to you.

where enabled, relay uploads relevant contact information to its systems.

you may revoke contact permission through your device settings.

revoking permission prevents future synchronisation but does not automatically delete previously uploaded contact information.

notifications

relay may request notification permission to provide:

  • live screening alerts;

  • urgent-call alerts;

  • completed-call notifications;

  • missed-call notifications;

  • call summaries;

  • service-status information;

  • security information; and

  • other relay-related updates.

notification preferences inside relay may operate independently from android’s system-level notification permission.

call handling and forwarding

relay may require call forwarding, telephony configuration, or other device or carrier settings to route eligible calls to the assistant.

disabling required call-handling functionality may prevent relay from screening calls.

15. how we use personal information

we may use personal information to:

  • provide relay;

  • authenticate users;

  • operate call routing;

  • screen calls;

  • transcribe speech;

  • generate responses;

  • generate summaries;

  • identify known callers;

  • apply your preferences;

  • provide notifications;

  • maintain call history;

  • provide customer support;

  • troubleshoot problems;

  • detect fraud or abuse;

  • maintain security;

  • measure service performance;

  • improve relay;

  • improve rumik’s ai and voice systems as described above;

  • process subscriptions;

  • comply with applicable law; and

  • establish, exercise, or defend legal claims.

16. how we share personal information

we do not sell or rent your personal information.

we may share personal information in the following circumstances.

service providers

we share information with providers where reasonably necessary to operate relay.

this may include providers involved in:

  • telephony;

  • realtime communications;

  • speech recognition;

  • ai inference;

  • voice generation;

  • cloud infrastructure;

  • authentication;

  • notifications;

  • analytics;

  • payments;

  • security; and

  • customer support.

communication with callers

when you instruct relay to communicate something to a caller, that information will necessarily be disclosed to that caller.

relay may also communicate contextual information reasonably necessary to handle a call on your behalf.

legal requirements

we may disclose information where reasonably necessary to:

  • comply with applicable law;

  • respond to valid legal process;

  • respond to lawful governmental requests;

  • investigate fraud;

  • investigate abuse;

  • protect users;

  • protect relay or rumik;

  • enforce applicable terms; or

  • establish, exercise, or defend legal claims.

business transfers

if rumik is involved in a merger, acquisition, financing, restructuring, sale of assets, or similar transaction, personal information may be transferred as part of that transaction subject to applicable law.

17. data retention

we do not retain every category of information for the same period.

17.1 raw call recordings

raw call recordings may be retained for up to 30 days.

during this period, they may be used for:

  • providing call history functionality;

  • durable transcription;

  • debugging;

  • customer support;

  • security investigations;

  • abuse investigations;

  • quality evaluation; and

  • permitted ai-improvement activities.

after the applicable retention period, recordings are deleted or irreversibly anonymised from systems under relay’s control, subject to:

  • temporary backups;

  • processor deletion schedules;

  • security requirements; and

  • legal obligations.

17.2 transcripts and summaries

transcripts, summaries, caller information, and related call-history information may remain associated with your relay account until:

  • you delete the call;

  • you delete your account;

  • the information is no longer required to provide relay; or

  • another applicable retention rule requires deletion.

17.3 uploaded contacts

uploaded contact information may remain associated with your account while relay provides caller-identification or related functionality.

after a valid account-deletion request, uploaded contacts will generally be deleted or irreversibly anonymised within 30 days, subject to applicable legal exceptions.

17.4 account deletion

when you submit a valid relay account-deletion request:

  • your account will be disabled;

  • uploaded contacts will be scheduled for deletion;

  • stored recordings will be scheduled for deletion;

  • transcripts will be scheduled for deletion;

  • summaries will be scheduled for deletion;

  • call history will be scheduled for deletion;

  • account preferences will be deleted or irreversibly anonymised; and

  • relevant processors will be instructed to delete information they process on our behalf where applicable.

we aim to complete deletion within 30 days of a valid request.

limited information may be retained where reasonably necessary for:

  • legal compliance;

  • accounting;

  • tax obligations;

  • fraud prevention;

  • security investigations;

  • dispute resolution; or

  • establishing, exercising, or defending legal claims.

information retained for these purposes will not be used for unrelated purposes.

17.5 backups

information may remain temporarily in encrypted backups after deletion from active systems.

backup copies are removed according to normal backup-expiration processes.

deleted information will not ordinarily be restored into active product systems except where necessary for disaster recovery, security, or legal compliance.

18. international data transfers

some service providers used by relay may process information outside india.

where personal information is transferred to another jurisdiction, we take reasonable measures designed to ensure that the transfer is permitted under applicable law and subject to appropriate technical, contractual, or organisational safeguards.

provider locations and infrastructure may change as relay develops.

where a material change affects how personal information is processed, we may update this privacy policy and provide additional notice where required.

19. data security

we use reasonable technical and organisational measures designed to protect personal information against:

  • unauthorised access;

  • loss;

  • misuse;

  • alteration;

  • disclosure; and

  • destruction.

these measures may include:

  • encryption in transit;

  • encryption at rest where appropriate;

  • authentication controls;

  • restricted access;

  • internal access controls;

  • access logging;

  • infrastructure monitoring;

  • secure development practices;

  • environment separation; and

  • security reviews.

no storage or transmission system can guarantee absolute security.

if a personal-data breach occurs, we will investigate the incident, take reasonable steps to contain and remediate it, and provide notifications where required by applicable law.

20. call recording and caller notice

telephone recording, transcription, interception, and notification requirements may differ based on location and circumstances.

relay may inform callers that:

  • they are communicating with an ai assistant;

  • relay is acting on behalf of another person;

  • the call or ai-handled portion may be recorded;

  • the conversation may be transcribed; and

  • information may be processed using ai.

relay users must not use relay for:

  • unlawful surveillance;

  • unlawful interception;

  • harassment;

  • fraud;

  • impersonation;

  • illegal data collection; or

  • other unlawful purposes.

21. consent and withdrawal

where processing relies on consent, relay may obtain consent before the relevant processing begins.

this may include consent relating to:

  • this privacy policy;

  • contact access;

  • call processing;

  • call recording;

  • transcription;

  • ai processing;

  • notifications;

  • third-party processing; and

  • other functionality requiring consent.

where applicable, you may withdraw consent by:

  • changing a relay setting;

  • revoking a device permission;

  • deleting your relay account; or

  • contacting us at [support-relay@rumik.ai](mailto\:support-relay@rumik.ai).

withdrawing consent does not affect processing that lawfully occurred before withdrawal.

certain relay functionality may stop working where the relevant data or permission is necessary to provide that feature.

22. your privacy rights

subject to applicable law, you may have rights regarding your personal information.

access

you may request information about personal data we process about you.

correction

you may request correction or completion of inaccurate or incomplete personal information.

deletion

you may request deletion of personal information where applicable.

valid deletion requests will generally be completed within 30 days, subject to applicable legal exceptions.

withdrawal of consent

where processing depends on consent, you may withdraw that consent.

grievance redressal

you may raise a complaint regarding how relay processes your personal information.

nomination

where available under applicable law, you may nominate another person to exercise applicable privacy rights in circumstances permitted by law.

to exercise a privacy right, contact:

[support-relay@rumik.ai](mailto\:support-relay@rumik.ai)

we may need to verify your identity before processing certain requests.

23. children’s privacy

relay is intended for account holders who are 18 years of age or older.

we do not knowingly allow children under 18 to create relay accounts.

if we learn that personal information belonging to a child has been collected in violation of applicable law or this policy, we will take appropriate steps to delete it.

a person calling a relay user may be under 18.

information communicated during an ordinary incoming call may necessarily be processed to handle that call, but relay does not intentionally create advertising profiles of child callers.

24. promotional and service communications

we may send service-related communications necessary to operate relay, including:

  • authentication messages;

  • security messages;

  • account information;

  • call-screening notifications;

  • subscription information;

  • important service notices; and

  • privacy-policy updates.

where permitted and where applicable, we may also send product announcements or promotional communications.

you may opt out of promotional communications using the method provided in the communication or through available relay settings.

service and security communications may still be sent where necessary to operate your account.

25. changes to this privacy policy

we may update this privacy policy as relay, rumik’s technology, service providers, product functionality, or applicable law changes.

when changes are material, we may notify you through:

  • the relay application;

  • email;

  • an in-app notice;

  • our website; or

  • another appropriate method.

where required by applicable law, we may request renewed consent before continuing the relevant processing.

the effective date at the beginning of this policy indicates when the policy was last updated.

26. grievances and privacy requests

for questions, complaints, data-access requests, correction requests, deletion requests, account deletion, consent withdrawal, or other privacy matters, contact:

relay privacy & support rumik intelligence private limited email: [support-relay@rumik.ai](mailto\:support-relay@rumik.ai)

we will address privacy grievances in accordance with applicable law.

where applicable under the dpdp act, you may also have the right to seek redress through the data protection board of india after using the grievance mechanism available through us.

27. contact us

for any questions about this privacy policy or relay’s data practices:

rumik intelligence private limited

relay is operated by rumik intelligence private limited.

join us

we are a close-knit group of researchers, engineers, and designers working on the hardest problems in ai.