INDUSTRIES · TELECOM
AI designed to keep subscriber data in India, with the record in your hands.
Care-floor assist, voice notes in Indian languages and corporate IT work, run inside your perimeter. Subscriber identifiers masked before a model reads them, and a recorded trail that helps when the six-hour clock starts.
IN SHORT
For telecom operators and service providers, AgentAnywhere is designed to run in India, inside your own perimeter. Seva assists live agents on the care floor and reviews quality, AgentAnywhere Vaani is designed to take voice notes in Indian languages on Indian infrastructure, Aavaran, our data-masking layer, masks subscriber identifiers before a model reads them, Manthan reads long documents and drafts for review, and Taksha works on software on your own network. Every step is recorded, in a trail you can export as CSV. This is IT-side and care-floor work, not network operation.
01 · Network data has a home
Under the 2026 authorisation rules, a new authorised entity keeps all data, logs and information associated with its telecommunication network within India.
02 · Six hours is short
The Telecom Cyber Security Rules, the CTI Rules and the CERT-In Directions each set a six-hour clock for incidents. A tool that touches the incident path must add to the timeline, not cloud it.
03 · Subscribers speak many languages
Care contacts arrive in Hindi, Tamil, Telugu, Marathi and English, often mixed in one sentence. They carry mobile numbers and ID numbers that should not reach a model in the clear.
WHY IT FITS
Sovereign, our models, and the Trust Platform.
Three things, in the terms of telecom: where it runs and who controls it, which models do the work, and the controls and the record around every call.
01 · Sovereignty
It runs in India, inside your own environment.
- Inside your perimeter: your own cloud account in an Indian region, or your own hardware. Region-pinned, and designed to deploy air-gapped.
- Swaraj, the sovereign deployment, is designed to pin a deployment to the region you choose, which suits data the rules keep in India.
- AgentAnywhere Vaani is designed to run on Indian infrastructure, so voice notes are handled in the country.
02 · Our models
The line we select, measure and serve, for the care floor and the IT team.
- Seva assists agents and reviews quality across Indian languages and English. AgentAnywhere Vaani handles voice notes in Indian languages; its masking, routing and tally are ours, and speech recognition and synthesis are open-licensed external components used for now.
- Manthan reads long operating documents and drafts for an engineer to check. Taksha works on billing and portal software on your network.
- Seva, Manthan and Taksha are built on open foundation models, measured and served by us. You can bring your own open model.
03 · The Trust Platform
Controls around every call, and a timeline you keep.
- Aavaran masks subscriber identifiers before a model reads them, with a signed Trust Receipt for each masking call.
- Kavach is designed to screen each request for prompt injection and unsafe output. The Agent Gateway applies your policy at the point of the call.
- Every step is recorded, in a trail you can query and export as CSV.
INSIDE YOUR PERIMETER
What happens to one subscriber's request.
Follow one request from the moment it arrives to the receipt an auditor can check. Raw values are designed not to cross the boundary.
One request · start to receipt
Illustration
Your perimeter · your cloud, your region, or designed to deploy air-gapped
Step 1 of 5
A request arrives
A question comes in. It carries a name, an account, a card number: the things that must not travel.
THE RULES
The rules that bind you.
What each instrument requires in plain words, where it stands today, and what it means for AI in telecom. Dates that have not arrived are written as dates.
Authorisation for Provision of Principal Telecommunication Services Rules 2026, G.S.R. 513(E), 23 June 2026, rule 30(3)(b)
Notified 23 Jun 2026
What it requires
A new authorised entity keeps all data, logs and information associated with its telecommunication network within India. The scope is network data of new authorised entities, not all subscriber data.
What it means for AI
An AI system that reads network logs for such an entity should run in India. Whether other data falls under the rule is for your counsel.
What it requires
The authorisation provisions of the Act, in force from 23 June 2026 (S.O. 3368(E)).
What it means for AI
The authorisation rules above attach data duties to the authorised entity. Plan AI tooling as part of that entity's systems.
Telecom Cyber Security Rules 2024, G.S.R. 720(E), 21 November 2024, amended by G.S.R. 771(E), 22 October 2025
Notified; amended Oct 2025
Text read on the DoT mirror, thc.nic.in
What it requires
Rule 7(1)(a): report an incident within six hours of becoming aware. Rule 6: a Chief Telecommunication Security Officer who is a citizen and resident of India.
What it means for AI
A recorded trail of what an AI tool read and did gives the security team a timeline inside the six hours. The report is yours to file.
Critical Telecommunication Infrastructure Rules 2024, G.S.R. 723(E), 22 November 2024
Only for networks notified as CTI
What it requires
For networks notified as CTI only. Rule 7(1)(j): notify incidents no later than six hours of occurrence. Rule 7(2): remote access from outside India needs prior written Government approval. Rule 7(1)(c): logs kept a minimum of two years.
What it means for AI
Remote access from outside India is what a hosted AI service or an overseas support session can mean. Keep the model and the tooling inside India, and scope any vendor access with this rule in the room.
What it requires
Report listed incidents within 6 hours of noticing. Keep ICT logs for a rolling 180 days within Indian jurisdiction.
What it means for AI
An AI system produces ICT logs. A recorded trail that stays in your deployment is the material for the log duty. Reporting stays with you.
What it requires
Commenced by G.S.R. 843(E) of 13 November 2025. Section 8(5): protect personal data, including data with a processor, by reasonable security safeguards. Rule 7(2)(b): a fuller breach report within 72 hours. Rule 6(1)(e): logs and personal data kept one year.
What it means for AI
Subscriber data is personal data. Masking before a model reads it, and a trail you can query, help you show the safeguards and rebuild an incident.
This is what the rules say, not legal advice. Your compliance function owns the reading and the filing.
USE CASES · BY FAMILY
What runs inside an operator.
Each card opens the page for that family or machine. Tell us your setting and we will show you the closest demonstration.
WORK SCENES
What it looks like on a working day.
The problem, what runs, and what the person at the desk gets. Each scene is an illustration of how the work is designed to go, not a customer account.
Scene 01
A mobile number read aloud on a call
The problem
- A subscriber reads out a mobile number and an ID number while an assist tool listens.
What runs
- Aavaran masks them in the transcript before the model reads it, and Seva suggests the next step.
What they get
- The agent sees a suggested reply and compliance flags, and the record shows a signed Trust Receipt for each masking call.
Scene 02
Voice notes that reach the IVR
The problem
- Callers who drop out of the IVR leave voice notes in many languages.
What runs
- AgentAnywhere Vaani is designed to understand, mask and classify each note on Indian infrastructure before an agent picks up.
What they get
- The agent starts with a masked, categorised request, and urgent or uncertain notes go to a person.
Scene 03
Quality on every contact
The problem
- A quality team reviews a sample of contacts, and the rest go unseen.
What runs
- Seva returns a summary and a quality scorecard with compliance flags for each interaction.
What they get
- The team leader opens the flagged contacts first.
Scene 04
A runbook question at night
The problem
- An on-call engineer needs the right page of a long runbook or vendor change note.
What runs
- Manthan reads the documents inside your perimeter and drafts an answer.
What they get
- The engineer reads the answer, checks it against the document and acts.
Scene 05
The incident review
The problem
- After an incident, the security team needs a timeline of what the AI tools read and did.
What runs
- Every step is recorded, and the trail can be queried by person, action and time.
What they get
- The team exports the trail as CSV. The six-hour report is yours to file.
THE SOVEREIGNTY TEST
The sovereignty test, for telecom.
If an AI system cannot answer these from its own records, you are renting it. Beside each question: what a rented AI service usually answers, and how we answer. Read the test.
01Where does the data run?
What a rented AI service usually answers
“In a region near you, per the contract.” Support staff may reach in from anywhere.
How we answer
In India, inside your perimeter, region-pinned and designed to deploy air-gapped. It runs where you run it, so you decide who can reach in, which matters for the CTI Rules on remote access.
02Which model made that decision?
What a rented AI service usually answers
“The AI.”
How we answer
Model Hub is designed to register each model with an owner and lineage. Seva, Manthan and Taksha are built on open foundation models, measured and served by us.
03What did the model actually see?
What a rented AI service usually answers
The full transcript, with the mobile number in it.
How we answer
Aavaran masks subscriber identifiers before the model reads them, with a signed Trust Receipt for each masking call.
04What stops the agent doing harm?
What a rented AI service usually answers
A prompt that tells the model to follow the rules.
How we answer
A policy outside the model. Kavach is designed to screen each request for prompt injection and unsafe output, and the Agent Gateway applies your policy at the point of the call.
05Who approved the consequential step?
What a rented AI service usually answers
“The system.”
How we answer
A person approves what goes out to a subscriber, and the approval is recorded.
06What will you show the regulator?
What a rented AI service usually answers
A slide for the regulator.
How we answer
A recorded trail, queryable and exportable as CSV, with a signed Trust Receipt for each masking call. Your compliance function owns the filing.
STATED PLAINLY
What this is not.
- This is IT-side and care-floor work. We do not operate, configure or monitor network elements.
- Rule 30(3)(b) covers network data of new authorised entities, not all subscriber data, and the CTI Rules apply only to networks notified as CTI. Your counsel decides how each applies to you.
- Seva is shown on synthetic conversations. AgentAnywhere Vaani is shown without accuracy figures, and its first pilot deployment is being built.
- Automated outbound voice is what the TRAI amendment of September 2026 describes as an A2P call. Our contact-centre work today assists a person on the call.
- Masking reduces what a model reads. It does not take data outside any law, and your compliance function owns that view.
THE OFFER · THE SOVEREIGN PILOT
Bring your hardest workload. We come to your data.
Pick the one workload your risk committee keeps deferring because the data cannot go anywhere. We run the pilot where the data already sits, with our architects beside your team, and you decide at the end whether it goes live.
For telecom operators: bring the care-floor workload with subscriber data in it. We set up in India, inside your perimeter, and the incident timeline is part of the record from the first call.
What the pilot is
- One workload, one team, six to ten weeks.
- A sovereign environment by default.
- Solutions Architecture paired with your team for the duration.
- It ends with one agent in production under your governance, or a documented decision not to proceed.
What we add
- The Trust Platform is on from the first call: data masking, policy screening, the Agent Gateway and a recorded trail.
- The record stays with you and is exportable as CSV.
- You can bring your own open model.
- A reply within one business day.
A risk committee has no reason to refuse, because the data does not have to go anywhere to find out. The pilot ends one of two ways: one agent in production under your governance, or a documented decision not to proceed, with the record behind it.
FAQ
Frequently asked questions.
- Can a telecom operator use AI without moving data out of India?
- Yes. AgentAnywhere is designed to run in India, inside your perimeter, in your cloud account in an Indian region or on your own hardware, region-pinned and designed to deploy air-gapped. Aavaran masks subscriber identifiers before a model reads them.
- What do the 2026 telecom authorisation rules say about data?
- Rule 30(3)(b) of the Authorisation for Provision of Principal Telecommunication Services Rules 2026, notified on 23 June 2026, says a new authorised entity keeps all data, logs and information associated with its telecommunication network within India. Its scope is network data of new authorised entities, not all subscriber data.
- How fast must a telecom incident be reported?
- Within six hours under three instruments: the Telecom Cyber Security Rules 2024 (rule 7(1)(a), from becoming aware), the Critical Telecommunication Infrastructure Rules 2024 (rule 7(1)(j), for notified networks) and the CERT-In Directions of 2022 (from noticing). A recorded trail of what an AI tool did gives your team a timeline. The report is yours to file.
- Can a vendor's staff access a critical telecom network from abroad?
- Under rule 7(2) of the Critical Telecommunication Infrastructure Rules 2024, remote access from outside India needs prior written Government approval. The rules apply only to networks notified as CTI. Keep that rule in the room when you scope any vendor's access, ours included.
- What can Seva do in a telecom care centre?
- Seva reads a service conversation and returns a summary, a quality scorecard, compliance flags and a suggested reply for a person to use. It is shown on synthetic conversations, and it is built on open foundation models, measured and served by us. Automated resolution is work we build with you.
- Does AgentAnywhere Vaani handle Indian languages?
- Seven to start: English, Hinglish, Hindi, Urdu, Tamil, Telugu and Marathi. Accuracy per language is being measured and is not yet published, and its first pilot deployment is being built. It is designed to run on Indian infrastructure.
- Does AgentAnywhere operate or monitor the network?
- No. This is IT-side and care-floor work: assist, quality review, voice notes, document reading and software on your own network. We do not operate, configure or monitor network elements.
- What does the Sovereign Pilot involve for an operator?
- One workload, one team, six to ten weeks, in a sovereign environment, with Solutions Architecture paired with your team. The Trust Platform is on from the first call, the record stays with you and is exportable as CSV, and it ends with one agent in production under your governance or a documented decision not to proceed.
Bring your hardest workload.
We come to your data. One workload, one team, six to ten weeks, in a sovereign environment. Tell us the workload and a solutions architect will reach out within one business day.