A welcome and onboarding AI agent integrates with the bank's core banking system (Finacle, BaNCS, T24, FLEXCUBE) and CRM to pull the new customer's product, profile, and acquisition channel data before the first call. The call is personalized by product type — a savings account welcome is entirely different from a credit card welcome, a personal loan welcome, or an insurance policy welcome — and by customer segment, language preference, and geography.

The agent's role spans six distinct onboarding functions: (1) warm welcome and expectation-setting — confirming what the customer has signed up for; (2) regulatory disclosure delivery — MITC for credit cards, KFS for loans, free-look period for insurance; (3) product activation — card PIN setup prompts, net banking first-login guidance, UPI registration walkthrough; (4) operational setup — NACH mandate for loan/insurance, salary credit instructions for salary accounts, nominee confirmation; (5) cross-sell introduction — contextually appropriate, not aggressive; (6) ongoing relationship setup — preferred contact channel, RM name and number, branch details.

The agent operates across a multi-touch onboarding journey rather than a single call. Industry data shows that a three-touch sequence (T+24h welcome call, T+7 activation follow-up if product not yet used, T+30 engagement review) drives 40–55% higher first-year product engagement compared to a single welcome SMS. The AI agent manages all three touchpoints automatically, with each triggered by CBS event data rather than a fixed schedule.

Welcome call timing follows the same 'speed to first contact' principle established in sales calling research: intent and engagement decay exponentially with time. A customer who just opened a digital savings account is excited about their decision, has the app in mind, and is primed to set up UPI and net banking. The same customer, called three days later, has moved on to other priorities and must be re-engaged from a lower baseline.

For credit card approvals, timing is even more critical. The card is approved but not yet activated. A customer who receives a welcome call within 2 hours of approval confirmation and gets guided through card activation completes that activation in the same session — with a first-transaction rate of 85–90% within 7 days. Customers who self-activate without a welcome call have a first-transaction rate of 60–70%, and 8–12% never activate their card at all.

For loan disbursement welcome calls, timing aligns with the disbursement event rather than the application approval. The optimal window is within 30–60 minutes of funds hitting the customer's account — when the customer is most engaged, most grateful, and most receptive to a conversation about repayment setup, prepayment options, and insurance coverage on the loan.

For insurance policies, IRDAI requires a policy welcome call within 15 days of issuance (aligned with the free-look period), but best practice is within 24–48 hours to ensure the customer understands their coverage before the free-look window is half spent. Kallix welcome agents trigger insurance calls within 4 hours of policy issuance.

A savings account welcome call is the most versatile onboarding flow because savings accounts serve as the gateway to the bank's full product suite. The agent's primary objective is to move the customer from 'account exists' to 'account is actively used' — the metric that determines long-term customer value.

The call opens with identity confirmation (date of birth or last four digits of mobile) and a warm welcome from the bank. The agent then guides the customer through the four activation steps that determine whether the account becomes active or dormant within 6 months: mobile app download (the agent sends the app store link via SMS during the call), net banking first-login (the agent walks the customer through the temporary password and first-login process), UPI ID registration (the customer's mobile number linked to the account becomes the UPI VPA), and debit card PIN setup (the agent explains whether the PIN is delivered by SMS, through the app, or via a green PIN letter).

Nominee confirmation is handled conversationally if the account was opened without a nominee — particularly important for accounts opened via video KYC or digital channels where nominee details are sometimes skipped. The agent explains why nominees are important (for deceased account claim processing) and collects the nominee's name, relation, and date of birth.

The final segment is a contextual product introduction: for a first-salary-earner, the agent introduces the recurring deposit or tax-saver FD; for a young professional opening their second bank account, the agent mentions the credit card offer availability; for a retired customer, the agent introduces senior citizen FD rates. Each introduction is one sentence — an offer, not a pitch.

RBI's Master Direction on Credit Cards (April 2022) mandates that banks deliver the Most Important Terms and Conditions (MITC) to new credit card holders at the time of card delivery. The welcome call fulfills this requirement verbally while the WhatsApp PDF provides the written record — satisfying both the spirit and letter of the disclosure obligation.

The MITC delivery segment covers the eight key terms in plain language, not fine print: 'Your credit limit is ₹[amount]. The interest rate on unpaid balances is 3.5% per month or 42% per annum. The minimum payment due is 5% of the outstanding balance or ₹500, whichever is higher. The late payment fee is ₹1,299 if you miss the minimum payment. Your annual fee is ₹[amount], waived if you spend ₹[amount] in the first year. Your billing cycle runs from the [X]th to the [X]th of each month. You have a 20-day grace period after your statement date to pay without interest. Cash advances attract an additional 2.5% fee and interest from day one — we recommend avoiding cash advances.' The agent then asks if the customer has questions on any of these terms.

Card activation is guided next: whether through the mobile app, IVR, SMS, or branch — the agent explains the specific activation method for that bank. First transaction incentive (welcome bonus, cashback on first ₹500 spend, or reward points activation) is mentioned with the specific offer details pulled from the CBS card record. The add-on card offer is made once, briefly, for the customer's spouse or family member. EMI facility on purchases above ₹5,000 is introduced in one sentence.

The KFS (Key Fact Statement), mandated by RBI for all retail loans under the RBI Digital Lending Guidelines 2022, is a standardized one-page summary of all loan terms. The welcome call's KFS delivery confirms that the customer understood what they signed — and gives the agent an opportunity to clarify terms that borrowers commonly misread, particularly the Annual Percentage Rate (APR) vs the quoted interest rate.

The KFS delivery covers: loan amount disbursed, interest rate and type (reducing balance vs flat), processing fee deducted, total amount payable over the loan tenure, monthly EMI amount and first EMI due date, prepayment policy (whether allowed after how many EMIs, whether charges apply), and foreclosure charges (typically 4–5% of outstanding principal for personal loans). The agent speaks these terms conversationally and then sends the KFS document via WhatsApp.

NACH mandate confirmation is the operationally critical step. The agent verifies that the mandate presented at loan application has been activated: 'Your EMI of ₹[amount] will be auto-debited from your [bank name] account ending [XXXX] on the [X]th of every month, starting [date]. Has your mandate been set up correctly?' If the customer reports their mandate was not submitted or has an issue, the agent immediately routes to the NACH setup sub-flow.

Insurance coverage advisory — credit shield, personal accident, or job loss cover on the loan — is introduced in the welcome call as a protection recommendation, not a hard sell. For customers who declined insurance at the loan application stage, the agent offers a 30-second explanation of what is covered before moving on.

NACH mandate setup is one of the most operationally valuable functions of a welcome call agent. A loan customer without an active NACH mandate is a credit risk management issue: the first EMI default rate for customers without NACH is 3–5x higher than for customers with active mandates, because manual payment requires active intent at each due date while NACH is automatic.

The e-NACH setup flow works as follows: the agent confirms the customer's repayment bank account (the account from which EMI should be debited) by having the customer provide the account number and IFSC. The agent verifies the IFSC is valid and the account is NACH-enabled via real-time API lookup. The customer then receives an OTP on their registered mobile from their bank (as the debit-account bank sends a verification OTP for NACH registration). The customer provides the OTP verbally; the agent submits the mandate to NPCI's NACH registration API. Mandate status is confirmed within 30–60 seconds.

For customers who want to use a different bank account for EMI repayment than the account used for the loan application, the agent handles the account change request and routes it to the loan operations team before completing the mandate setup.

Where e-NACH cannot be completed during the call (customer doesn't have the repayment account OTP, or the bank is not NACH-enabled), the agent schedules a callback for the next 2 hours and sends a WhatsApp reminder with the steps the customer should keep ready — bank account number, IFSC, and mobile number registered with the repayment bank. The callback converts at 55–65% to completed mandate within 24 hours.

Digital banking activation within the first week is the single strongest predictor of customer lifetime value. A customer who makes their first digital transaction within 7 days of account opening has a 4–6x higher 12-month product engagement rate than one who makes their first digital transaction after 30 days. The welcome call's guided activation is the primary mechanism driving this outcome.

The agent sends the mobile app download link via SMS during the call — the customer sees the SMS arrive, taps the link, and downloads the app while still on the phone. The agent then guides first-login: temporary login ID format (usually customer ID or mobile number), one-time password delivery method (SMS or email), and how to set a permanent password that meets the bank's security requirements. This live guidance eliminates the 'forgot my password' and 'invalid credentials' failure modes that cause 40–50% of self-serve first-login abandonment.

UPI registration is guided next, either through the bank's own UPI feature in the mobile app or through BHIM, PhonePe, or Google Pay. The agent explains the UPI VPA format (mobilenumber@bankname) and confirms the UPI handle setup by asking the customer to send a ₹1 test transfer to a known number — converting UPI registration from an abstract setup task to a completed action.

For customers who cannot complete the app download during the call (on a feature phone, in a low-connectivity area, or not comfortable with app installation), the agent provides the net banking URL, the first-login credentials method, and schedules a follow-up call for digital activation assistance.

IRDAI regulations mandate a 15-day free-look period for all life insurance policies (30 days for policies sold through distance marketing channels), during which the customer can return the policy and receive a full refund minus proportionate risk premium and stamp duty. The welcome call serves two functions with respect to free-look: ensuring the customer knows the right exists (reducing post-period cancellation disputes) and giving the customer confidence in their purchase (reducing free-look period cancellations by informed customers who were simply uncertain).

The call covers policy-type-specific content. For term life insurance: sum assured, policy term, annual premium amount, nominee details, premium payment due date, and grace period (typically 30 days). For health insurance: sum insured, whether it's floater or individual, waiting periods for pre-existing conditions (typically 2–4 years), exclusions category summary, cashless hospital network query assistance, and how to raise a claim. For ULIP/endowment products: the distinction between insurance coverage and investment component, premium allocation, and surrender charges.

The agent explicitly states the free-look period terms: 'You have 15 days from today — that is until [date] — to review your policy in detail. If you decide this policy is not right for you, you can return it to us and receive a full refund minus a small proportionate risk premium for the days covered. Do you want me to explain how to exercise this option?' This direct disclosure — rather than hoping the customer reads the policy document — is a significant trust-building action and satisfies IRDAI's suitability and disclosure requirements.

Nominee confirmation is an important secondary function: policies sold through bancassurance channels sometimes have missing or incorrect nominee details. The agent verifies the nominee name, relationship, and date of birth during the welcome call.

First-30-day engagement is the foundational metric for customer lifetime value in retail banking. Industry data consistently shows that customers who make at least three product interactions in the first 30 days have a 12-month retention rate of 85–92%, compared to 55–65% for customers with zero or one interaction. The AI welcome calling journey is designed to engineer those three interactions.

Touch 1 (T+24h welcome call): product activation, regulatory disclosure delivery, NACH setup, digital banking guidance, and a warm RM introduction. This call sets the customer's expectation of the bank relationship and resolves the most common first-week friction points before they become complaints.

Touch 2 (T+7 activation follow-up): triggered only for customers who have not made a digital transaction or product interaction within 7 days of the welcome call — a CBS inactivity event drives the trigger. The follow-up call is brief (3–4 minutes), focused exclusively on unblocking the specific friction: 'I noticed you haven't yet set up net banking — did you have any difficulty with the first-login process? Let me guide you through it right now.' This targeted intervention converts 45–55% of initially inactive new customers into active users within the following 7 days.

Touch 3 (T+30 engagement review): a structured review call covering product usage confirmation, unresolved service issues from the first 30 days, and a contextual next-product introduction. By day 30, the agent has CBS data on the customer's transaction pattern — and the product introduction is informed by that data: a customer who has been making 15+ monthly transactions is introduced to a credit card; a customer with a consistent monthly salary credit is introduced to a home loan pre-qualification; a customer with growing savings is introduced to FD or RD rates.

Personalization is the key differentiator between a welcome call that feels like a genuine relationship opener and one that feels like a robocall. The AI agent pulls the customer's profile from CBS and CRM before the call: product type, account type, acquisition channel, city, language preference, age, and any available credit profile data.

Age and life stage drive substantial variation. A first-time bank account opener aged 22 needs mobile app guidance and UPI setup as the primary activation — they will naturally migrate to digital. A senior citizen opening a fixed deposit account needs a different flow: slower pace, emphasis on RM contact for branch assistance, nominee setup, and FD maturity instruction options rather than digital channel push. The AI agent detects age from the date of birth in CBS and adjusts pacing and content accordingly.

Acquisition channel matters significantly. A customer acquired through an aggregator portal (BankBazaar, PaisaBazaar) often has lower bank brand familiarity than a branch-acquired customer — the welcome call needs a stronger brand introduction and clearer explanation of what bank they've signed up with. A DSA-acquired loan customer may have had a salesperson over-promise product features — the welcome call's KFS delivery is an opportunity to reset accurate expectations before the first EMI.

Credit profile segmentation affects the cross-sell logic in the Day-30 call. A thin-file customer (credit bureau score below 650, or no credit history) should not receive a credit card offer — instead, the agent introduces the secured credit card (card against fixed deposit) or a credit-builder product. An established credit customer (CIBIL 750+) can be offered pre-approved products with specific offer details pulled from the bank's pre-approval engine.

Unanswered calls are a normal occurrence in welcome calling — new customers may have an unfamiliar number calling, may be busy at work, or may be in a no-call zone. The cascade strategy maximizes reach without becoming intrusive.

The first attempt at T+2h post-account opening is the primary window. The second attempt at T+6h (for morning sign-ups, this means an early-evening call; for afternoon sign-ups, this means a next-morning call) is timed to catch a different behavioral window for the same customer. The third attempt at T+24h is the final voice outreach. The caller ID shows the bank's registered service name — not an unknown number — which improves answer rates significantly compared to generic outbound numbers.

After three unanswered voice attempts, the agent sends a WhatsApp message with the welcome summary content: account number confirmation, key activation steps with app store links, MITC/KFS PDF attachment (for credit card and loan customers), and a callback request button. WhatsApp messages achieve 85–92% delivery and 60–70% read rates, making this an effective fallback for voice-unreachable customers.

For customers in DND-registered numbers: TRAI TCCCPR 2018 classifies welcome calls as transactional communications (not commercial), exempting them from the Do-Not-Disturb registry. However, the bank must use its registered service provider name and call only between 9 AM–8 PM. All Kallix welcome calls operate within these parameters.

For customers who request 'don't call me,' the agent respects the preference immediately, sends the welcome summary via WhatsApp, and flags the customer's CRM record as 'voice opt-out' — future onboarding touchpoints are delivered via WhatsApp and email only.

The welcome call cross-sell philosophy is anchored in the principle that adjacency sells — a customer who just opened a savings account is a natural prospect for a debit-linked credit card, not a mortgage; a customer who just took a personal loan is a natural prospect for loan protection insurance, not a fixed deposit. Contextual adjacency creates relevance; relevance creates openness.

For savings account welcome calls: the adjacent cross-sell is a debit-card-linked credit card offer (if the customer is pre-approved based on their credit profile), an FD for the initial account funding amount, or a recurring deposit set up for the amount the customer mentioned as monthly savings. These are offered in the final two minutes of the call as a natural extension of the setup conversation.

For credit card welcome calls: the adjacent cross-sell is a balance transfer offer (for customers who mention other bank cards), personal accident insurance bundled with the card, or an EMI conversion offer on their first large purchase. The add-on card for a family member is both a product and a usage activation strategy — add-on card issuance drives the primary cardholder's spend by 35–45%.

For loan welcome calls: the adjacent cross-sell is loan protection insurance (critical illness, job loss, or accidental death cover on the outstanding principal), a savings account for salary credit if the customer doesn't bank with the lender, or a top-up loan after 6–12 months of clean repayment.

For insurance welcome calls: the adjacent cross-sell is a complementary policy (term + health as a natural pair; health + personal accident as a natural pair). The agent introduces the complementary product in a 'complete your protection coverage' framing after the free-look advisory.

Incomplete KYC is common for digitally acquired accounts — the account is provisionally opened but operational limits apply until full KYC is completed. An account with incomplete KYC typically faces credit limits of ₹1 lakh per year on inflows, no outbound transfers above ₹10,000, and no card issuance. The welcome call is the ideal moment to surface and resolve this, before the customer tries to use the account and hits an unexplained restriction.

The agent's CBS lookup at call initiation flags KYC completeness. If any KYC fields are outstanding, the agent covers them explicitly after the welcome segment: 'I can see that your account is fully active for most transactions, but we need one more document to remove the transaction limits — a recent utility bill or rental agreement as address proof. You can upload this directly in the app, or I can schedule a short video verification call with you in the next 48 hours. Which would you prefer?'

For Video KYC (V-CIP) completion, the agent schedules the 10–15 minute video session by routing to the appointment booking sub-flow — integrating with the calendar system to find a slot where a video KYC agent is available. A confirmation WhatsApp with the video link and required documents list is sent immediately.

For FATCA declarations (mandatory for potential US person accounts), the agent explains the regulatory requirement in plain language and routes the customer to the digital FATCA form via a link sent on WhatsApp. For accounts where FATCA is unresolved beyond 30 days, RBI requires the bank to restrict account operations — the agent explains this consequence clearly and creates urgency without creating alarm.

Demat and investment account welcome calls operate within SEBI's investor protection framework. The account welcome is both a functional onboarding (how to use the platform) and a regulatory obligation (ensure the customer understands the product they've signed up for, particularly risk profile).

The Demat account welcome covers: the 16-digit Beneficiary Owner (BO) ID (also called the Demat account number), the Depository Participant (DP) ID (the broker's ID), the depository (CDSL or NSDL — which determines the Demat statement access portal), how to access the Demat statement, and how to set up a linked bank account for settlement (T+1 settlement cycle explanation).

For new mutual fund account holders, the welcome call introduces the SIP (Systematic Investment Plan) setup process: amount, date, fund selection guidance (not advice — the agent explains what information is available in the platform for the customer to choose), and mandate setup for SIP auto-debit. The NACH mandate setup for SIP functions identically to the loan NACH sub-flow — the agent completes it in-call if the customer is ready, or schedules a follow-up.

For first-time equity investors, SEBI's investor protection guidelines require disclosure of the risk nature of equity investments. The agent delivers a 30-second risk disclosure: 'Investments in securities are subject to market risk. Past performance of any market instrument does not guarantee future returns. Please review your risk profile on the platform and ensure your investment choices align with your risk tolerance.' This is logged with a timestamp as the investor education disclosure.

KYC and KRA (KYC Registration Agency) status is confirmed — the customer's KYC should be registered with one of SEBI's KRAs (CDSL Ventures, NDML, CAMS, Karvy). If KRA status is pending, the agent advises the timeline and next steps.

Early churn — accounts opened but never actively used — is a substantial cost problem for retail banks. A dormant account incurs CBS maintenance cost, statement delivery cost, and regulatory reporting cost with zero revenue contribution. The 90-day dormancy rate for digitally acquired accounts at most Indian retail banks is 18–25%, representing a meaningful waste of customer acquisition cost.

The primary drivers of early churn are: activation friction (customer couldn't complete app setup or card activation), unmet expectations (the product is different from what the customer expected based on the sales conversation), and life circumstance changes (the customer found a competing product, the DSA gave them a better offer elsewhere). Welcome calling addresses all three.

For activation friction: the T+7 inactivity call identifies specifically which activation step the customer failed at — the agent asks directly, identifies the blocker, and guides resolution in the same call. For credit cards, this converts 45–55% of inactive T+7 cardholders into active users. For savings accounts, this converts 40–50% of non-digital-banking T+7 customers into app users.

For unmet expectations: the KFS/MITC delivery in the T+24h welcome call resets any over-promises made during the acquisition phase. Customers who feel the product was accurately described are 60% less likely to close the account within 90 days than those who feel they were misled. The welcome call is a trust investment — its ROI is measured in retention, not in activation numbers.

For competitive churn: the T+30 engagement review call identifies customers who have opened a competing product and are drifting. A targeted retention offer — waiver of annual fee, enhanced FD rate, pre-approved credit product — made at the T+30 call converts 25–35% of at-risk customers back to primary bank status.

RM handoff is a critical quality gate in the welcome journey. For high-value customers, the experience of being welcomed by an AI agent without an RM follow-up can feel underwhelming — the relationship expectation is higher. The AI agent is configured to identify high-value customers from CBS data at call initiation and route them immediately after the regulatory disclosure segment to a human RM.

The warm transfer mechanism works as follows: the AI agent tells the customer 'I'm going to connect you with your dedicated relationship manager, [Name], who will continue this conversation' and simultaneously sends the RM a call brief via the bank's CRM — customer name, product, acquisition channel, what was covered so far in the call, and any questions or concerns the customer raised. The RM picks up the call with full context rather than a cold introduction.

For customers who express dissatisfaction during the welcome call — 'I was told the annual fee was waived', 'the agent who sold me this said the interest rate would be lower' — the AI agent does not attempt to resolve these complaints itself. It acknowledges the concern, documents it as a complaint reference, and routes to either a complaints resolution specialist or the customer's RM. This is a protective escalation: attempting to dispute a sales promise during a welcome call without authority to resolve it damages trust further.

The AI agent's CRM integration ensures the RM has full visibility of all welcome journey touchpoints before any human interaction. When the Day-30 engagement review identifies a customer ready for a product upgrade or credit increase conversation, the agent either books an appointment with the RM directly or, for pre-approved digital decisions, completes the conversation itself.

NRI customers represent one of the highest-value customer segments for Indian banks — inward remittances average ₹8–₹15 lakh per year per NRI household, and their propensity to hold fixed deposits, insurance, and investment products is higher than resident customers. The welcome call must reflect this value while addressing the specific practical and regulatory context of the NRI banking relationship.

Time zone alignment is the first operational consideration. NRI welcome calls are scheduled based on the customer's declared country of residence at account opening, with a fallback to the declared preferred contact window in the CRM. Calls to Gulf region customers (UAE, Saudi, Qatar — largest NRI banking segment in India) are scheduled for early morning IST to align with Gulf evening. Calls to US/Canada customers are made late evening IST.

Remittance setup is the primary activation task for NRI savings accounts. The agent provides: the NRE or NRO account IBAN, the bank's SWIFT code (BIC), the correspondent bank details for the receiving country (the intermediary bank), and instructions for marking the remittance purpose code (FEMA purpose code for inward transfers, e.g., P0008 for family maintenance). This information converts the account from an abstract account number to a functional receiving account.

For NRE vs NRO distinction, the agent explains the two key differences: NRE account income is fully repatriable and exempt from Indian income tax; NRO account income from Indian sources is taxable in India and repatriation requires a CA certificate up to $1 million per year. The agent does not provide specific tax advice but explains the distinction accurately enough for the customer to make informed product use decisions.

Salary accounts are acquired through corporate relationships — the bank has a tie-up with the employer, and the employee opens an account under that relationship. The welcome call must reflect both the individual employee's needs and the corporate relationship context.

The most operationally important task is salary credit setup. Even when the bank has a corporate tie-up, many employees have existing salary accounts at other banks and need to update their employer's HR or finance team with the new account details. The agent confirms the account number and IFSC, advises the customer to submit a written request to their HR or payroll team within 5 business days for the current month's salary to credit correctly, and sends the account details via WhatsApp for the customer to forward.

Salary-linked benefits vary by the corporate tie-up and the employee's salary grade. Common benefits the agent covers: zero balance maintenance (the account has no minimum balance requirement as long as salary credits are regular), accidental insurance cover (typically ₹2–₹5 lakh, automatic with the account), a pre-approved credit card offer based on the declared salary, and an overdraft facility of 1–2 months' net salary. The agent activates or explains activation steps for each benefit present in the customer's product profile.

The salary lapse advisory is important: most salary accounts convert to a regular savings account (with minimum balance requirements) if no salary credit is received for 3 consecutive months — for example, if the employee changes jobs and doesn't update the new employer's payroll with this account. The agent explains this proactively so the customer is not surprised by a minimum balance charge after a job change.

Measuring welcome call effectiveness requires going beyond call completion and contact rate — the downstream behavioral metrics are what determine whether the welcome journey is working. A welcome call with a 90% contact rate but a 40% 7-day activation rate indicates that the call is reaching customers but not completing the activation objective. Kallix's reporting framework separates process metrics (contact rate, call duration) from outcome metrics (activation, usage, retention).

The six KPIs and their benchmarks: Activation rate (% of customers who complete primary product activation within 7 days): target 72–80% for credit cards, 80–88% for savings account digital activation. Digital adoption rate (first digital transaction within 7 days): target 85–90% for credit cards, 75–82% for savings accounts. NACH mandate setup rate: target 65–75% for loan customers. 30-day product usage rate (at least one transaction in first 30 days): target 88–92%. 90-day churn rate (accounts closed or dormant within 90 days): target under 14%. NPS of welcome call: target 55–65, measured via post-call SMS survey.

KPIs are tracked at the product level and the acquisition channel level — because branch-acquired customers have different baseline rates than digitally acquired customers. A credit card welcome call for a digitally acquired customer will have different activation dynamics than one for a customer who applied through a bank branch.

Monthly performance reports compare welcome call cohorts (customers who received welcome call) against control cohorts (customers who received only SMS welcome) to isolate the causal impact of the AI calling program. These reports drive script optimization and call timing adjustments.

System integration is the backbone of a personalized welcome call. Without CBS integration, the agent cannot verify account status, KYC completeness, product details, or NACH mandate status in real time. Without CRM integration, the agent cannot personalize by acquisition channel, language preference, or credit profile segment. Without calendar integration, the agent cannot book RM appointments. Without WhatsApp/SMS integration, post-call document delivery cannot be automated.

CBS integration at call initiation pulls: account number, account type, product details, KYC status fields, NACH mandate status (for loan/insurance), credit limit (for cards), first transaction status, and acquisition channel code. This data is refreshed at call start — the agent does not work from a static pre-generated report.

CRM integration pulls: customer language preference, acquisition date, acquiring RM name and contact (for warm transfer setup), any open service requests or complaints (the agent checks for unresolved issues before delivering a welcome — a customer with an unresolved complaint needs resolution before an activation pitch), and pre-approval eligibility flags from the credit engine.

For Salesforce FSC deployments, the agent creates or updates a Contact record and creates a new Interaction record for the welcome call, with call outcome, activation status, and cross-sell interest flags populated automatically. The RM sees the complete welcome journey in the customer's Salesforce timeline before their first human interaction.

For Leadsquared deployments (common in NBFCs and new-age banks), the agent updates the lead stage from 'Disbursed/Issued' to 'Active Customer' upon successful activation completion, triggering the appropriate nurture workflow in the CRM.

Home loans are the highest-value and longest-tenure banking relationship — a 20-year home loan customer has a total interest outflow of ₹40–₹80 lakh on a ₹50 lakh loan. The welcome call for a home loan disbursement is not just an activation call — it is the foundation of a two-decade relationship. The agent's role is to ensure a smooth start, resolve any first-week anxieties, and introduce the support ecosystem.

Disbursement confirmation is the opening segment: the agent confirms the tranche amount, the account it was credited to (for underconstruction properties, disbursements are often sent directly to the developer's account, not the customer's), the remaining sanctioned amount to be disbursed in future tranches, and the expected disbursement timeline for remaining tranches.

For under-construction properties, the agent confirms whether the customer's NACH-linked pre-EMI interest schedule has started. Pre-EMI interest (interest on the disbursed portion before full disbursement and EMI commencement) is a common source of confusion — customers expect the EMI to start after possession, but pre-EMI interest begins immediately upon each disbursement. The agent explains this clearly and confirms the pre-EMI amount and payment method.

Property insurance advisory covers two distinct products: mortgage reducing term assurance (MRTA — a single-premium policy that covers the loan outstanding if the borrower dies), and home structure insurance (fire and allied perils insurance for the property itself, often required by the bank as a loan condition). The agent explains both, confirms whether either was bundled with the loan, and offers to connect the customer with the bank's insurance team if interested.

Tax benefit advisory is one sentence: 'You can claim deductions up to ₹1.5 lakh under Section 80C for principal repayment and up to ₹2 lakh under Section 24B for interest paid in a financial year — your tax advisor can help you optimize this based on your income bracket.'

Welcome calls involve personal data processing, mandatory product disclosures, outbound call communication, and in some cases financial advice — each requiring specific compliance treatment.

Under TRAI TCCCPR 2018, welcome calls are transactional communications — they inform the customer about their own account or product, which exempts them from the Do-Not-Disturb registry restrictions. However, the bank must use its registered service provider name as the caller ID, must call only between 9 AM and 8 PM, and must honor opt-out requests immediately. The agent's voice opt-out handling flags the CRM record immediately to prevent future calls.

Under DPDP Act 2023, call recording requires explicit consent at the start of the welcome call. The agent delivers a two-sentence consent notice: 'This call may be recorded for training and quality purposes. Your product and account information will be used to personalize this conversation and improve our services to you. Do you consent?' Consent is logged with a timestamp. Welcome call recordings are retained per the bank's customer interaction data policy, typically 2 years.

RBI MITC delivery for credit cards (Master Direction, April 2022) requires verbal delivery at time of card issuance or delivery. The welcome call fulfills this obligation, with the call recording serving as the audit trail for compliance review. The WhatsApp PDF delivery provides the written confirmation.

For RBI KFS under Digital Lending Guidelines 2022, the KFS must be provided before loan disbursement. The welcome call's KFS delivery is a post-disbursement confirmation — the pre-disbursement KFS delivery happens in the loan approval workflow. The welcome call KFS is logged separately as post-disbursement confirmation.

IRDAI's Grievance Redressal Guidelines require insurance policy welcome calls to include explicit free-look period advisory. The agent's free-look disclosure with the specific date and cancellation process is logged as part of the policy servicing record.

Account dormancy — defined by RBI as a savings account with no customer-initiated transaction for 24 months — is a significant operational and regulatory issue. While the immediate concern is 90-day dormancy (an early warning indicator), the longer tail is the 24-month RBI dormancy classification that triggers inoperative account status and eventual transfer to RBI's Depositor Education and Awareness Fund (DEAF) after 10 years.

The AI agent's dormancy prevention call at T+60 is data-driven: the trigger fires only for customers whose CBS transaction data shows zero customer-initiated transactions (not bank-initiated fees or standing credits) in the first 60 days, and who did not complete the digital banking activation in the welcome journey. These customers are identified by a CBS query rather than time-based scheduling, ensuring the call is targeted and not wasted on already-active customers.

The call opens with a specific observation rather than a generic engagement push: 'I can see you've had your [account type] with us for about two months. I wanted to check in to see if you've had a chance to explore the account and if there's anything that has been unclear or difficult.' This framing is respectful and diagnostic — it positions the call as support rather than sales.

The engagement incentive is personalized: for a savings account, a cashback on the first digital transaction (₹50 cashback on first UPI payment, for example) or an enhanced FD rate for the idle balance in the account. For a dormant credit card, the incentive might be a welcome bonus reactivation (where the first-use bonus hasn't been triggered because the card was never used) or a 10% cashback on the first ₹1,000 spend.

Current account welcome calls serve a business customer whose banking needs are operationally more complex than a retail savings account. The agent's objective is to get the business account operationally functional for outward payments within the first week — the most urgent need for most businesses opening a new bank account.

Business internet banking setup is the primary activation task: the agent confirms the company's authorized signatory details, guides the corporate login setup (which differs from retail — corporate internet banking often requires maker-checker workflows, transaction limits per signatory, and multi-factor authentication tied to DSC or hardware tokens), and sends the first-login instructions to the email address registered for the account.

Account operator configuration covers: who can initiate transactions (makers), who can approve them (checkers), and what the transaction limits are for each operator role. For SMEs with a single authorized signatory, this is simple. For mid-market companies with treasury teams, the agent explains the multi-signatory configuration process and routes to the business banking operations team for setup.

GST-linked transaction features (specific to Indian current accounts): many banks offer integrated GST payment options where the current account is directly linked to the GST portal for tax payment, and automated GST-related transaction tagging in account statements. The agent introduces this feature and routes to the business banking team for setup.

Currently Account welcome calls end with a specific RM introduction: the business banking RM's name, mobile number, and email are provided verbally and via WhatsApp — the business customer needs a named human contact, not a generic customer service number.

Premium and HNI customers have a heightened service expectation that makes a fully AI-handled welcome call inappropriate for the entire onboarding journey. However, the AI agent plays a valuable role in the opening segment: it ensures regulatory disclosures are delivered accurately and completely, creates a clean compliance audit trail, and handles the initial identity confirmation — freeing the RM from a scripted opening and allowing them to focus on relationship-building from the start.

The agent's HNI welcome opening covers: confirmation of account or product details (takes 30 seconds), MITC or KFS delivery if applicable (takes 2–3 minutes), and a specific statement setting up the RM transfer: 'Your dedicated relationship manager at [bank name] is [RM Name], who handles your portfolio personally. I'm connecting you to [RM Name] now — they have full details of your account and are expecting your call.' The agent then executes the warm transfer with the CRM brief.

For private banking and wealth management welcome calls, the RM is typically pre-briefed via the CRM before the AI agent initiates the call — the AI agent's role is more of a concierge handoff than a standalone call. The AI initiates the call, confirms identity, delivers any regulatory segment (investment risk disclosure, AIF KYC confirmation), and transfers to the private banking RM.

For corporate salary accounts with premium benefit packages (expense management, corporate travel insurance, Lounge access, business card), the AI agent handles the benefits orientation call while routing any specific business banking advisory to the corporate relationship manager. The benefits orientation is structured and can be fully AI-handled — it is informational rather than relational.

The Day-30 review is the most data-rich call in the onboarding journey because by day 28, the CBS has 4 weeks of actual transaction behavior for the customer — real intent signals rather than the assumed profile used at day 1. The agent uses this data to make the call genuinely relevant rather than generically welcome.

The call opens with a personalized observation: 'I can see you've been using your account actively — you've made [X] transactions this month including at [merchant category]. That's great to see.' This confirmation of observed behavior is a powerful engagement signal — it shows the customer the bank is paying attention to their relationship, not just holding their money.

The unresolved service issue sweep is a structured two-question segment: 'Have you had any issues or questions about your account in the first month that weren't resolved?' and 'Is there anything about your account that doesn't work the way you expected?' These open questions surface latent complaints before they become NPS detractors or churn events. Day-30 complaints identified and resolved have a significantly higher satisfaction recovery rate than complaints left to escalate.

The next-product introduction uses the CBS transaction pattern as the data source. A customer making frequent small UPI payments at merchants is a natural credit card candidate (cashback on small transactions). A customer who received a large fund transfer from overseas is an NRI remittance product prospect. A customer with a large idle balance for 30 days is an FD or systematic deposit prospect. The agent introduces the single most relevant product — one product, one sentence offer — and books an appointment with the RM if the customer expresses interest.

Welcome calling ROI is multi-dimensional — it is not simply a cost-per-call comparison. The full value calculation must account for churn prevention, cross-sell revenue, credit quality improvement, and compliance risk avoidance.

Cost comparison: a manual welcome call team handling T+24h, T+7, and T+30 touchpoints for 5,000 new customers per month requires 10–15 dedicated agents working a full shift, with loaded cost of ₹400–₹600 per customer for the three-touch sequence. AI handles the same 5,000 customers for ₹80–₹140 per customer — a cost reduction of ₹1.3–₹2.3 crore annually for a mid-sized NBFC or bank branch network.

Churn reduction value: if a bank acquires new customers at a cost of ₹1,200–₹2,000 per customer (CAC including marketing, DSA commission, and onboarding cost), and welcome calling reduces 90-day churn by 25–35%, then the churn reduction saves ₹300–₹700 per customer in recovered CAC. For 5,000 monthly new customers, this is ₹1.5–₹3.5 crore annually in protected acquisition investment.

Cross-sell revenue: at 8–15% cross-sell conversion from Day-30 calls, for 5,000 monthly customers with an average second-product revenue of ₹2,000–₹3,000 per year, the cross-sell program generates ₹96 lakh–₹2.25 crore in incremental annual revenue.

NACH setup improvement: a 25–35 percentage point improvement in NACH setup rates (from 30–40% to 65–75%) reduces first-EMI default risk. For an NBFC with ₹100 crore monthly disbursements, even a 0.5% reduction in first-EMI default rate is worth ₹50 lakh annually in reduced NPA provisioning.

Welcome calling is faster to deploy than dispute or fraud agents because the CBS integration is read-only (no write access required), there is no card network API configuration, and the compliance framework is simpler — MITC delivery and consent logging rather than dispute investigation audit trails.

Week 1–2 covers CBS integration (read-only customer and account data pull), CRM integration (customer profile, segment, and RM data), and WhatsApp/SMS gateway setup. The CBS integration is typically an API connection to the bank's existing customer data layer rather than a direct CBS API — reducing information security review time.

Week 3 covers call flow scripting: distinct welcome scripts for each product type the bank offers (savings, credit card, personal loan, insurance, demat, home loan, salary account, current account). Kallix provides pre-built templates for all major Indian banking product types. Customization covers the bank's specific MITC terms, KFS parameters, RM routing rules, and cross-sell offer set.

Week 4 covers compliance review: the bank's compliance team reviews MITC and KFS delivery scripts, DPDP consent language, free-look period advisory text, and TRAI-compliant caller ID configuration. Typically the lightest compliance review in the Kallix deployment suite — 3–5 days.

Week 5 is the pilot: first 500–1,000 new customers across two or three product types, with Kallix call quality specialists monitoring all calls in real time. Adjustments based on pilot call listening typically require 1–3 days. Full rollout to all new customers follows in week 6.

The case for AI welcome calling is not that AI is better than humans at building relationships — it is that manual welcome calling at scale is operationally difficult to do consistently and quickly, and most of its content is structured enough to be automated without losing quality.

Script adherence is the most significant quality difference. Mandatory regulatory disclosures — MITC, KFS, free-look period — must be delivered completely and verbatim to every customer. Human agents under call volume pressure skip or abbreviate disclosures at a rate of 18–25% per call quality audit across Indian banking call centres. AI agents deliver 100% of mandatory disclosures on every call — the compliance value alone often justifies deployment.

Timing is the second difference. A human welcome call team calling 500 new customers in a day has a practical throughput limit — calls queue and the first-call timing slips from the ideal T+2–4 hours to T+12–24 hours or beyond. AI has no throughput constraint — the T+2 hour call is made for every customer simultaneously, regardless of daily volume.

Consistency of tone and handling is the third difference. A human agent handling their 80th welcome call of the day sounds different from a fresh agent handling their third — tiredness, impatience, and script fatigue affect the customer experience measurably. AI delivers the same energy and completeness on call 1,000 as on call 1.

Where human agents genuinely outperform AI in welcome contexts: open-ended advisory conversations ('what product should I take?'), empathy-intensive interactions with distressed or confused customers, and high-value relationship conversations with HNI customers. The Kallix model routes all three to humans.

New customer churn concentrates in the first 90 days — and the leading cause is an unresolved onboarding problem the customer never had a chance to report. Without a proactive welcome call, these customers silently stop using the product, reduce balance, and eventually close the account or let it go dormant.

Kallix welcome call complaint discovery protocol:

1. Check-in question: after confirming the account activation and product features, the agent asks: 'Before we finish — is there anything about your account opening experience or account access that is not working the way you expected?' This open-ended question is the highest-yield complaint discovery method — 68% of customers who have an issue will mention it when asked directly, versus 12% who proactively call the helpline.

2. Complaint classification: the agent classifies the issue into one of four categories — (a) technical: app login, card activation, feature not visible; (b) documentation: wrong name, wrong address on welcome kit, PAN mismatch; (c) service: long branch wait, staff behaviour; (d) product: unexpected charge, feature not as described.

3. In-call resolution: technical issues with known fixes (app cache clear, PIN reset, feature toggle) are resolved by the agent in the same call. Time to resolution for in-call resolvable issues: under 4 minutes.

4. Warm transfer: for issues requiring back-office or branch action, the agent creates a service request with priority flag (new customer within 30 days), provides the SR number to the customer, and warm-transfers to a senior RM or complaint handling team — not a general queue.

5. Follow-up: a follow-up call is scheduled within 48 hours to confirm resolution. If unresolved at 48 hours, the SR is escalated to the branch head.

Lenders using Kallix complaint discovery on welcome calls see 42% lower 90-day churn among customers who surface an issue versus customers with unresolved friction who never had a complaint channel offered proactively.