Health insurance benefit inquiries are the highest-volume servicing call type for insurers — policyholders want to know exactly what is covered before a hospitalization event, not during it. An AI voice agent handles the full call arc: identity verification (policy number + DOB, with OTP fallback for shared-family policies), real-time PAS query, benefit recitation in plain language, and structured confirmation.

The agent reads sub-limits in rupees, not percentages: 'Your room rent is capped at Rs 5,000 per day for a general ward and Rs 10,000 for a single A/C room. ICU charges are covered in full.' This removes the ambiguity that drives post-claim disputes. For family floater policies, the agent confirms the aggregate sum insured, how much has been utilized in the policy year, and the remaining available balance per member.

For corporate group health policies, the agent queries the HR portal API or group master data to confirm employee-specific sum insured (base + top-up), co-payment (if any), and whether spouse and children are enrolled. Termination-date grace periods (typically 30 days for group policies) are confirmed when the employee queries coverage post-resignation.

Kallix integrates with Majesco, DuckCreek, FinanceFriend, and legacy AS400-based PAS systems via REST, SOAP, or SFTP batch. TPA APIs (Medi Assist, MDIndia, Paramount, Vipul Medcorp) are queried in real time for network status and benefit limits. Integration typically completes in 4–6 weeks.

Waiting period inquiries are disproportionately high-stakes because a caller who gets the wrong answer may either delay a necessary admission or face a claim repudiation they weren't expecting. AI agents eliminate human error by computing remaining months algorithmically from the policy inception date and the insurer's waiting period schedule.

For PED waiting periods, IRDAI Health Insurance Regulations 2016 mandate a maximum of 48 months (4 years) for any pre-existing condition. Most insurers apply 24 months for standard health policies and 48 months for senior citizen plans. The agent confirms: 'Your diabetes was declared at enrollment. The 36-month waiting period ends on 14 March 2026. Any hospitalization for diabetes before that date will not be covered under your current policy.'

For maternity, waiting periods range from 9 months (Apollo Munich/HDFC Ergo select plans) to 4 years (entry-level plans). The agent also checks: (a) whether the sum insured includes a maternity sub-limit (Rs 50,000–Rs 2,00,000 is typical), (b) newborn coverage — IRDAI mandates day-1 newborn coverage under the IRDAI Health Insurance Regulations amendment 2019, (c) whether C-section is covered at the same limit as normal delivery.

For specific illness waiting periods (e.g., hernia, cataract, joint replacement — typically 24 months under standard plans), the agent queries the product-level waiting period matrix in the PAS. For portability cases, the agent confirms the crediting of completed waiting period from the previous insurer under IRDAI portability norms.

Cashless hospital lookup is the second-highest-volume query for health insurance servicing teams — and one of the most error-prone when handled manually, because hospital empanelment lists change frequently as TPAs add or remove networks. AI agents with live TPA API connectivity always return current empanelment status, avoiding the common problem of policyholders arriving at a hospital that has been de-empanelled.

The lookup flow: caller gives pin code (or city + locality), agent queries TPA network API filtered by insurer product code and specialty (cardiac, orthopaedic, oncology, maternity, dialysis, etc.), then ranks results by distance. For NABH-accredited hospitals — which receive 5–10% higher reimbursement under some insurer tariffs — the agent flags this as financially relevant to the caller.

For network-sensitive situations: (a) day care procedures — the agent filters for hospitals specifically empanelled for the day care procedure code; (b) critical illness — the agent can filter for NABH Tier 3 hospitals with the relevant specialty; (c) mental health — following IRDAI/Mental Healthcare Act 2017 mandate, the agent confirms whether the insurer's network includes NIMHANS-affiliated or Vandrevala-network mental health facilities.

For corporate group health policies with insurer-specific network restrictions (e.g., employer has negotiated cashless only at preferred network hospitals), the agent applies the group product code filter before querying. This prevents the scenario where an employee uses a general TPA network hospital that is outside their employer scheme's approved list.

TPA routing confusion is a persistent pain point: policyholders call the insurance company, the insurance company redirects them to the TPA, and the TPA asks for information the policyholder doesn't have at hand. AI agents collapse this loop by (a) instantly identifying the TPA from the policy record, (b) providing the relevant TPA helpline (pre-auth desk, cashless desk, claims helpline), and (c) initiating the TPA pre-auth notification if the hospitalization is planned.

The four major TPAs in India — Medi Assist (35 million+ lives), MDIndia, Paramount Health Services, Vipul Medcorp — each have documented API endpoints for pre-auth notification. Kallix integrates with these APIs to push the pre-auth notification with: policy number, proposed date of admission, hospital code, diagnosis code (ICD-10), and expected duration. This triggers the TPA's pre-auth workflow and reduces the average pre-auth turnaround from 4–6 hours to 90 minutes (planned procedure data, Medi Assist 2023).

For emergency pre-auth (admission already occurred), the agent follows the 24-hour retrospective notification pathway — flagging that the insurer must be informed within 24 hours of admission under standard policy conditions, and initiating that notification immediately via API. For policyholders who have already been admitted and are calling mid-hospitalization, the agent confirms the pre-auth reference number from the TPA API and reads out the approved amount and procedure scope.

For corporate group health, the agent routes to the dedicated corporate desk of the TPA (not the retail helpline), which has faster response times and access to the employer scheme's benefit schedule.

Medical inflation in India runs at 14–17% per annum (Mercer Marsh Benefits India Healthcare Trend Report 2024), making sum insured adequacy a genuine financial risk for families who bought their health policy 5+ years ago. An AI agent that surfaces this risk proactively — not just when the caller asks — delivers measurable value and creates a natural cross-sell opportunity.

The adequacy check logic: the agent takes the policyholder's city tier (Tier 1: Mumbai, Delhi, Bengaluru; Tier 2: Pune, Hyderabad, Chennai; Tier 3: smaller cities), family composition (number of members, eldest age), and sum insured, then benchmarks against the insurer's claims data for that city-tier and age band. It then states the gap specifically: 'For a family of four in Bengaluru with the eldest member aged 52, our 2025 claims data suggests Rs 7–10 lakh sum insured as adequate. Your current cover is Rs 3 lakh.'

The recommended solution is typically a super top-up policy — which triggers at a deductible (e.g., Rs 3 lakh deductible aligned to existing base cover, super top-up of Rs 10–20 lakh). Super top-ups are significantly cheaper than top-line base cover: a Rs 10 lakh super top-up with Rs 3 lakh deductible for a 45-year-old costs approximately Rs 8,000–12,000 per annum versus Rs 22,000–28,000 for a fresh Rs 10 lakh base cover.

For group health policyholders with employer-provided cover, the agent highlights the portability gap: employer cover ends at resignation, and buying a fresh individual policy at age 40+ triggers fresh waiting periods. The agent recommends purchasing a top-up or individual plan now to avoid waiting period re-initiation.

Claims history statements are requested for several purposes: income tax filing (Section 80D premium deduction documentation), employer reimbursement claims, fresh insurance application at a new insurer (disclosure requirement), and portability applications. An AI agent that handles this end-to-end — query, recitation, and dispatch — eliminates the 3–5 day manual process common in insurer back offices.

For tax filing, the agent confirms: (a) total premium paid in the financial year (for 80D deduction claim), (b) claims paid by the insurer in the year (which do not affect the premium deduction eligibility), and (c) the tax receipt if the policyholder needs it separately. Section 80D allows deduction of Rs 25,000 for self/spouse/children and an additional Rs 50,000 for parents aged 60+, making this a high-frequency query in January–March every year.

For portability applications, the claims history is critical because IRDAI portability norms require disclosure of all claims in the prior 3 years. The agent confirms the claims count and total settled amount, and pre-populates the portability form fields where the insurer's system supports API-based portability submission.

For group health, where employees may have had multiple employers and multiple TPAs, the agent focuses on the current policy period and routes multi-insurer consolidation queries to a human advisor. The PDF statement generated includes the insurer's stamp, TPA reference numbers, and a digital signature via Leegality or eMudhra for regulatory admissibility.

E-health card requests spike at policy renewal (new policy year, new card needed), after endorsements (new member added, new card needed), and when policyholders lose or misplace existing cards ahead of a hospitalization. Each of these scenarios is time-sensitive, making AI-powered same-session dispatch a significant improvement over manual queues.

The card generation flow: agent confirms identity (policy number + registered mobile OTP), queries PAS for all insured members under the policy, confirms the member(s) needing the card, calls the TPA API to generate the e-health card PDF (with QR code linked to the TPA's verification system, policy details, member photo if on file, and network hospital QR shortcut), then sends via email and WhatsApp simultaneously.

For group health policies, e-health cards are batch-issued by the insurer/TPA at group renewal, but individual employees frequently call mid-year for re-issuance after losing their card or adding a dependent. The agent handles single-member re-issuance from the group master without requiring HR involvement, which is a significant servicing improvement for employees in the field.

For senior policyholders (65+) who may struggle with digital cards on smartphones, the agent offers to read out the key card details — TPA name, helpline, policy number, and insured member names — as an immediate alternative while the digital card is dispatched to a family member's contact.

OPD coverage is one of the most misunderstood health insurance features in India because: (a) most base health plans do not include OPD, (b) those that do have significant sub-limits and category restrictions, and (c) corporate group health plans increasingly include OPD wallets managed through digital platforms like Eka Care, MFine, or DocsApp — which operate differently from traditional insurer OPD reimbursement.

For policies with OPD cover, the agent confirms: the annual OPD limit (Rs 5,000–Rs 20,000 for standard plans; Rs 50,000–Rs 1,00,000 for premium plans like Niva Bupa ReAssure or Aditya Birla Activ Health Platinum), the eligible expense categories (general consultation, specialist consultation, diagnostics, pharmacy, dental, vision — each may have sub-sub-limits), the claim submission process (original bills or scanned copies accepted), and the reset date (policy anniversary for most plans, April 1 for group health).

For corporate OPD wallets linked to telehealth platforms, the agent confirms the enrolled platform (Eka Care, MFine, DocsApp, 1mg), the wallet balance, how to book a consultation, and how pharmacy orders are processed. This requires API integration with the platform rather than the insurer PAS, but Kallix supports this through webhook-based data exchange.

For Ayushman Bharat PM-JAY beneficiaries, the agent confirms that PM-JAY covers only in-patient hospitalization (no OPD) but checks eligibility for state-specific OPD schemes (e.g., Karnataka Arogya Sanjeevani, Tamil Nadu CMCHIS) where applicable.

Critical illness riders are sold as a lump-sum benefit payable on first diagnosis of a listed condition — but policyholders frequently don't understand the survival clause, the definition of 'first diagnosis,' or whether their specific diagnosis qualifies under the rider's medical definitions. AI agents trained on the product's technical document and benefit definitions can resolve these inquiries definitively.

Standard critical illness riders in India cover 10–12 conditions at the base level: cancer of specified severity, first heart attack of specified severity, stroke, kidney failure requiring dialysis, coronary artery bypass surgery, major organ transplant, aorta graft surgery, primary pulmonary arterial hypertension, and blindness. Premium riders (e.g., HDFC Life Critical Illness Plus, Max Life Critical Illness and Disability) cover 36–64 conditions including early-stage cancers, multiple sclerosis, Parkinson's, and motor neurone disease.

The survival period clause — most riders require the policyholder to survive 30 days post-diagnosis — is frequently misunderstood by grieving families calling on behalf of a deceased policyholder. The AI agent handles this with the appropriate empathy protocol: acknowledges the situation, explains the 30-day survival clause factually, and immediately routes to a life claims specialist if the survival period was not met, as some riders have a special terminal illness or death benefit fallback.

For the claim initiation flow: the agent registers the claim in the CMS, dispatches the document checklist (histopathology report or cardiac catheterization report as applicable, hospital discharge summary, investigation reports confirming the diagnosis, attending specialist certificate), and schedules a 48-hour follow-up call to confirm document receipt.

Mental health coverage inquiries are a rapidly growing category — IRDAI data shows a 340% increase in mental health claims between 2019 and 2024, driven by post-COVID mental health awareness, IRDAI's 2018 mandate, and insurers' gradual network expansion for psychiatric facilities. Many policyholders are still unaware that their policy legally must cover mental health hospitalization.

The IRDAI mandate (September 2018) requires insurers to cover: in-patient psychiatric hospitalization at par with physical illness hospitalization, day care procedures for psychiatric treatment (e.g., electroconvulsive therapy, intensive outpatient programs), and de-addiction treatment for alcohol/substance use disorder. Insurers cannot apply sub-limits below the standard hospitalization benefit for mental illness.

The AI agent's response flow: (1) confirms IRDAI mandate exists and applies to the policy, (2) reads any rider-specific mental health enhancements from the PAS, (3) queries the TPA network for empanelled psychiatric facilities in the caller's city — filtering for NABH-accredited facilities and those with NIMHANS affiliation where relevant, (4) for corporate group health, confirms whether the employer's benefit schedule includes telehealth mental health consultations (increasingly common in tech sector group health plans).

For de-addiction treatment inquiries, the agent confirms that NRHM-recognised de-addiction centres are covered under the hospitalization benefit, confirms the maximum duration of covered stay (typically 30 days per episode under standard plans), and flags the pre-auth requirement. The agent does not stigmatise or apply different tone — the same transactional, factual approach used for physical illness applies.

HIV/AIDS coverage was a significant gap in Indian health insurance until IRDAI's July 2019 circular. Policies issued before October 2019 may still exclude HIV/AIDS under legacy exclusion clauses; policies issued or renewed after that date must include coverage. This creates a two-tier population that requires careful handling in the AI agent's response logic.

The AI agent's response protocol: (1) checks the policy issuance/renewal date against the October 2019 effective date, (2) if the policy is post-October 2019, confirms HIV/AIDS coverage and reads the standard hospitalization benefit — no separate limit, no discriminatory sub-limit, IRDAI-mandated parity, (3) if the policy is pre-October 2019 and has not been renewed since, flags that a renewal would activate IRDAI-mandated HIV/AIDS coverage, (4) confirms the 30-day waiting period for HIV-related admissions (maximum permitted under IRDAI circular).

For treatment-specific queries — ARV (antiretroviral) therapy, opportunistic infection treatment, AIDS-defining illness hospitalization — the agent confirms these are covered under the standard hospitalization benefit. Day care procedures for HIV management (e.g., IV administration of ARV drugs) are covered under the 900+ IRDAI-mandated day care procedures list.

The agent handles these queries with the same factual, non-stigmatising tone as any other clinical inquiry, as required by the Mental Healthcare Act 2017's non-discrimination provisions and the insurer's internal dignity guidelines. Routing to a human agent is offered proactively for complex multi-drug interaction queries.

Day care procedure inquiries are common because policyholders scheduled for procedures like cataract surgery, chemotherapy infusion, dialysis, or endoscopy are unsure whether the same-day procedure qualifies for cashless or reimbursement under their policy. The AI agent resolves this without the caller needing to visit a hospital to find out.

The IRDAI day care list (mandated under IRDAI Health Insurance Regulations 2016, Circular 2016) includes procedures such as: cataract surgery, chemotherapy infusion, dialysis, angiography, endoscopy (upper/lower GI), ESWL (kidney stone lithotripsy), hydrocele repair, tonsillectomy, adenoidectomy, varicose vein ligation, hysteroscopy, joint aspiration, and 880+ other coded procedures. No insurer can exclude these by applying a 24-hour minimum stay clause.

For each procedure inquiry, the agent: (1) confirms the procedure is on the mandated day care list, (2) queries the TPA network for empanelled day care centres (some procedures can only be done in specific empanelled facilities — e.g., lithotripsy at certified urological centres, chemotherapy at oncology day care centres), (3) initiates the pre-auth notification with the procedure code, estimated date, and chosen facility, (4) confirms the pre-auth approval timeline (typically 30–60 minutes for standard day care procedures, 2–4 hours for oncology).

The agent also confirms post-procedure follow-up claim requirements: for reimbursement, day care requires the same document set as in-patient (discharge summary, procedure notes, pharmacy bills), with the additional note that 'discharge summary' for day care will show same-day admission and discharge.

Pre- and post-hospitalization coverage is systematically under-claimed in India — an estimated 60–70% of eligible policyholders do not file pre/post claims, leaving significant money on the table. AI agents that proactively inform policyholders of this benefit at claim registration or policy inquiry calls directly increase claim utilization and policyholder satisfaction.

Pre-hospitalization coverage (30 days): covers diagnostic tests (blood work, imaging, ECG, echocardiogram), specialist consultations leading to the admission decision, and prescribed medications taken prior to admission — all directly related to the hospitalization diagnosis. The agent confirms that the cause link is required: pre-hospitalization expenses for cardiac tests before a cardiac bypass are covered; unrelated expenses in the same 30-day window are not.

Post-hospitalization coverage (60 days): covers follow-up consultations, physiotherapy, prescribed medications, and diagnostics (e.g., post-operative blood tests, imaging to confirm recovery) for 60 days after the discharge date. Some premium plans extend this to 90 or 180 days. The agent reads the specific post-hospitalization period from the PAS and confirms the claim submission deadline.

Claim filing flow for pre/post: the agent explains that pre/post claims are filed separately from the in-patient claim — either as part of the same reimbursement bundle (preferred) or as a separate post-discharge submission. Original bills with treating doctor's prescription linkage are required. For cashless hospitalizations, pre/post claims are always reimbursement (not cashless) — the agent proactively informs policyholders of this distinction to avoid post-discharge confusion.

Family floater utilization is a frequent inquiry — especially mid-policy-year when one family member has already been hospitalized and others want to know how much cover remains. AI agents with live TPA API access can answer this precisely rather than directing callers to wait 3–5 days for a statement.

The utilization query flow: agent authenticates, queries the TPA claims ledger for the policy year, and reads: total sum insured (e.g., Rs 5 lakh), total claims paid to date (e.g., Rs 2.1 lakh against Member B's appendectomy), remaining balance (Rs 2.9 lakh), and pending claim amounts under adjudication. If a claim is under review, the agent flags: 'Member B has a claim of Rs 85,000 currently under assessment — if approved, your remaining balance would be Rs 2.05 lakh.'

For families where the floater is substantially depleted mid-year, the agent proactively recommends a super top-up with a deductible equal to the floater sum insured — so if the floater is exhausted, the top-up activates. This recommendation is timed to be most effective: within 14 days of a large claim settlement when the depletion is visible.

For renewal: the agent confirms that the sum insured is reinstated in full at the next policy anniversary — explaining the reset date and cautioning that any claims currently under adjudication (not yet settled) will count against the prior year's balance and will be deducted from the reinstated sum insured if they settle post-renewal.

PM-JAY (Pradhan Mantri Jan Arogya Yojana) covers approximately 107 million beneficiary households (over 500 million individuals) from the bottom 40% of India's economic distribution — the largest government health scheme in the world by coverage. Millions of eligible beneficiaries do not know they are eligible or how to access cashless treatment. AI voice agents deployed by insurance companies, CSCs (Common Service Centres), or NBFC-MFIs can bridge this awareness gap at scale.

The eligibility check flow: the agent takes the Aadhaar number (masked, UIDAI-compliant) or ration card number, queries the NHA Beneficiary Identification System (BIS) API, and returns: eligible/not-eligible status, number of covered family members, and scheme name (PM-JAY directly or state scheme — e.g., Mahatma Jyotirao Phule Jan Arogya Yojana in Maharashtra, Aarogyasri in Andhra Pradesh/Telangana, CM Comprehensive Health Insurance Scheme in Tamil Nadu).

For eligible beneficiaries, the agent confirms: (a) Rs 5 lakh per family per year cover, (b) covers 3 days pre-hospitalization and 15 days post-hospitalization, (c) cashless at PM-JAY empanelled government and private hospitals — the agent queries the NHA hospital locator API for the nearest empanelled hospital, (d) no premium to be paid by beneficiary — scheme fully funded by government, (e) e-health card can be generated via Ayushman App or CSC.

For beneficiaries who hold commercial health insurance in addition to PM-JAY (increasingly common as micro-insurance penetrates the BPL segment), the agent explains that commercial insurance is the primary cover — PM-JAY activates after commercial coverage is exhausted, and dual-coverage coordination with the TPA is required.

Policy document re-issuance is a high-volume servicing request — particularly after renewal (new policy period document needed), after endorsements (updated schedule), and when policyholders misplace documents ahead of a hospitalization. AI agents with PAS document generation API access eliminate the 3–7 day cycle for physical policy kit delivery.

The document generation flow: agent authenticates, confirms the email and WhatsApp number on record (or updates to new contact if required), triggers the policy document API to generate a current policy schedule PDF, and sends simultaneously. For IRDAI compliance: the policy document dispatched digitally must carry an electronic signature (DSC/eSign via Leegality or eMudhra) to be legally equivalent to a physical policy. Kallix integrates with both providers.

For KYC and contact detail updates: IRDAI Digital Insurance Regulations 2023 permit V-CIP (video-based KYC) for insurance policy servicing — the agent schedules a V-CIP session via the insurer's V-CIP provider (Signzy, IDfy, or CAMS), which completes in 3–5 minutes via video call. CKYC update is then pushed to CERSAI's central KYC registry. Address updates trigger a zone re-check for motor insurance (zone A/B/C) and cashless network re-mapping for health insurance.

For group health, KYC updates for individual employees (name change after marriage, address change after relocation) are routed through the HR portal API to maintain the group master data integrity — direct policyholder updates without HR approval are flagged for group master reconciliation rather than processed immediately.

Room rent sub-limits cause disproportionate claim disputes because of the proportionate deduction clause: if a policyholder takes a room costing Rs 8,000/day when the policy cap is Rs 5,000/day, the insurer applies a proportionate deduction to all charges — not just the room rent. A 40% room rent excess may trigger a 40% proportionate cut on surgery fees, anaesthesia, nursing, and ICU charges, converting an apparently small upgrade decision into a Rs 1–2 lakh out-of-pocket shock.

The AI agent's pre-hospitalization advisory: 'Your policy covers room rent up to Rs 5,000 per day for a general ward. If you take a room above this limit, all hospital charges will be proportionately reduced. For your planned procedure, we recommend staying within the Rs 5,000 room rent limit unless your treating doctor specifies a medical need for a higher-category room.' Medical necessity (doctor certificate) is a valid basis to waive the proportionate clause at some insurers — the agent flags this option.

For ICU sub-limits: most standard policies cover ICU at 2× the room rent limit (e.g., Rs 10,000/day if room rent is Rs 5,000). Premium policies cover ICU in full. The agent reads the specific ICU limit from the PAS and confirms whether the limit is daily or aggregate. For critical care requiring extended ICU stays (5–10 days), the agent proactively calculates the policyholder's potential ICU co-payment exposure if the stay extends beyond the daily limit.

Policies without any room rent sub-limit (e.g., Niva Bupa ReAssure, HDFC Ergo Optima Restore) are confirmed as such — 'Your policy has no room rent restriction; you may choose any room category without proportionate deduction.' This distinction is a significant upgrade selling point and is used in the renewal reminder call.

Health insurance portability was introduced by IRDAI in 2011 and is significantly under-utilized — less than 2% of health policyholders port annually, despite many having legitimate reasons (insurer service quality, premium increase, network inadequacy). AI agents can proactively identify portability candidates and guide them through the process, either to retain them (by resolving their underlying concern) or to support their portability to a Kallix-served insurer.

The portability eligibility check: (a) the policy must be individual health or family floater (group health portability has different rules), (b) the policyholder must have been continuously insured for at least one year without lapse, (c) portability application must reach the new insurer at least 45 days before the current policy's renewal date (IRDAI regulation; failure to meet this deadline forfeits the portability cycle for that year).

What transfers under portability: (a) waiting period credit — years completed for PED and specific illness waiting periods at the current insurer carry forward; the new insurer must grant credit for completed years, (b) NCB/no-claim benefit — the accumulated no-claim bonus on the sum insured transfers, (c) continuity benefit — pre-existing conditions that have crossed the waiting period continue to be covered without re-initiation of waiting period.

What does not transfer: (a) the current insurer's network hospitals (policyholder must check the new insurer's network), (b) product-specific benefits (e.g., recharge/refill benefit — new insurer may not have an equivalent), (c) loyalty discounts from the current insurer. The agent presents a structured comparison and recommends the portability only if the benefit-to-premium change is clearly in the policyholder's favour.

Corporate group health insurance servicing is a distinct workflow from individual retail health — employees typically don't know their TPA, their benefit limits, whether their dependants are enrolled, or which hospitals are covered under the employer scheme (which may differ from the TPA's general network due to employer-specific panel agreements).

Employee authentication uses employee ID + registered mobile OTP, with fallback to date-of-joining and date-of-birth for employees who haven't updated their mobile. The agent then reads the grade-based benefit schedule: base sum insured, whether the employer has provided a parental cover (top-up for parents), whether there is an employee-paid top-up option, and the co-payment (if any — typically 0% for IT/ITES, 10–20% for manufacturing).

Dependent management: mid-year addition of a newborn or spouse requires the agent to initiate a group endorsement request to the employer's HR portal. The agent collects the dependent's details, generates a pre-filled endorsement form, and routes it to HR for approval — rather than requiring the employee to navigate the HR portal manually. Removal of a dependent (e.g., post-divorce or child over age limit) follows the same flow.

Claim query routing: for pre-auth, the agent contacts the TPA corporate desk (which has the employer scheme schedule loaded) rather than the retail helpline. For claim status, the agent queries the TPA's corporate claims dashboard. For GMER (Group Medical Expenses Reimbursement), the agent guides on document submission and expected reimbursement timeline (typically 15–21 days for complete documentation).

Co-payment disputes are the second-largest category of health insurance IGMS complaints after room rent. The core problem: policyholders misunderstand whether co-payment applies to all claims or only specific scenarios, and they are blindsided at discharge. AI agents that proactively explain co-payment before hospitalization — not after — convert a potential complaint into a prepared policyholder.

Co-payment structures in India vary significantly: (a) flat co-payment (e.g., 10% for all claims — common in senior citizen plans), (b) senior citizen co-payment triggered by age (e.g., 10% above age 60 for general plans that add a senior rider), (c) specific disease co-payment (e.g., 20% for joint replacement at specified hospitals), (d) zone-based co-payment (e.g., 20% co-payment if policyholder is treated outside their city of residence — relevant for corporate policyholders who travel), (e) room rent excess co-payment (the proportionate deduction triggered by exceeding room rent limit — which functions economically like a co-payment but is different in mechanism).

Deductible (versus co-payment): a deductible is a fixed amount the policyholder pays before insurance activates (e.g., Rs 50,000 deductible on a super top-up — the base cover pays the first Rs 50,000, top-up pays beyond). The agent explains the deductible in the context of the policyholder's claim scenario and confirms whether the deductible is aggregate (per year) or per-claim.

For IRDAI compliance: IRDAI Health Insurance Regulations 2016 prohibit co-payment for senior citizens above 65 who have been continuously insured with the same insurer — a protection that many insurers apply silently without informing policyholders. The agent flags this entitlement when the policyholder is 65+ with 3+ years of continuous coverage.

Health insurance NCB and refill benefits are among the most valued but least understood features of modern health policies, and policyholders consistently underutilize them because they don't know what they've accumulated. AI agents that read and communicate these benefits proactively — not just when asked — increase policyholder satisfaction and perceived value.

Health NCB (distinct from motor NCB): in health insurance, NCB increases the sum insured each claim-free year rather than reducing the premium. Star Health, HDFC Ergo, and Niva Bupa plans offer 5–10% sum insured increase per claim-free year, up to 50–100% total accumulation. For example, a policyholder who started with Rs 5 lakh 6 years ago with no claims may now have an effective Rs 7.5–8 lakh sum insured via accumulated NCB. The agent reads this number directly from the PAS and states it explicitly: 'Your accumulated no-claim bonus has increased your effective sum insured to Rs 7.5 lakh.'

Refill/recharge benefit: increasingly standard in premium health plans (HDFC Ergo Optima Restore, Niva Bupa ReAssure, Aditya Birla Activ Health), refill restores the base sum insured once per year after it has been exhausted by a claim — effectively doubling the annual coverage for Rs additional premium. The agent confirms: whether the plan includes refill, whether the refill has already been activated in the current year, what amount was restored, and whether the restored amount can be used for the same illness as the original claim (the key restriction: Optima Restore excludes the same illness, ReAssure allows same illness after 45 days — the agent reads the product-specific rule).

For policyholders considering renewal, accumulated NCB is a key retention argument: porting or letting the policy lapse resets the NCB to zero, with significant rupee-value consequences on coverage.

Health insurance AI voice servicing is only as good as the data it can access — a system without real-time PAS and TPA connectivity can only answer generic policy questions, not the specific, actionable queries policyholders need answered before a hospitalization. Kallix's integration architecture covers the full servicing data stack.

PAS integration: (a) Majesco Health — REST API with OAuth 2.0 authentication; supports policy read, endorsement initiation, document generation; (b) DuckCreek OnDemand — OData API endpoints for policy and claims; (c) FinanceFriend — REST + MySQL/MSSQL direct read for read-only queries; (d) AS400/legacy — SFTP batch file exchange with 15-minute refresh for benefit and policy data. New PAS platforms are onboarded via Kallix's integration layer with a typical 3–4 week connector build.

TPA integration: (a) Medi Assist — REST API for pre-auth notification, claim status, network hospital lookup, and e-health card generation; (b) MDIndia — REST API for pre-auth and claim status; (c) Paramount Health Services and Vipul Medcorp — API integration with standardised HL7 FHIR R4 endpoints for clinical data exchange (diagnosis codes, procedure codes, pre-auth scope);

Data security: all PAS/TPA API calls traverse Kallix's SOC 2 Type II compliant infrastructure with field-level encryption for Aadhaar, health data, and claim amounts. Health data handling complies with IRDAI's data localisation requirements and aligns with DPDP Act 2023 consent framework for health data as a sensitive personal category. No health data is stored beyond the session context unless explicitly required for follow-up workflows.

For insurers with no live API: Kallix supports a screen-scraping + RPA fallback for legacy web portals, typically used for pre-auth status lookup from insurer back-office web systems that predate API availability. This fallback has a 3–5 second higher latency than the API path but covers 98%+ of Indian insurers.

Maternity benefit inquiries are highly time-sensitive — expectant mothers calling during the third trimester need precise, actionable answers, not generic policy language. AI agents trained on product-level maternity schedules provide the specificity required.

Maternity sub-limits in India's standard health market range from Rs 50,000 (normal) / Rs 75,000 (C-section) on entry-level plans to Rs 1,00,000 / Rs 1,25,000 on mid-tier plans to Rs 2,00,000 / Rs 2,50,000 on premium plans (Niva Bupa Maternity Plus, HDFC Ergo Optima). The agent reads the specific sub-limits from the PAS and states both normal and C-section limits explicitly — because the treating doctor's decision on delivery mode is made at a medical basis, not financial, but the policyholder should know the coverage difference in advance.

Newborn coverage: IRDAI Health Insurance Regulations 2019 amendment mandates day-1 newborn coverage under the parent's family floater policy. The agent confirms: (a) the newborn is automatically covered from birth under the existing floater for 90 days without endorsement, (b) after 90 days, a formal endorsement request must be initiated to permanently add the newborn to the policy (the agent schedules a reminder call at 75 days), (c) the newborn's coverage under the floater is subject to the aggregate floater sum insured (not a separate sum insured for the newborn).

Maternity-related add-ons: some premium plans include maternity benefit extensions for complications of delivery, NICU coverage for newborn ICU stay (Rs 50,000–Rs 1,50,000), and infertility treatment (IVF) — typically under a specific rider with a separate waiting period of 2–4 years. The agent confirms whether these riders are active and the applicable sub-limits.

Grievance handling is a high-stakes servicing category because mishandled or unacknowledged grievances escalate to IGMS or Insurance Ombudsman complaints, both of which trigger IRDAI monitoring and reputational risk for the insurer. AI agents that register grievances correctly, confirm the reference number in real time, and set accurate resolution expectations prevent 35–50% of IGMS escalations (production data, Kallix insurance deployments).

Grievance registration flow: the agent logs the grievance type (claim rejection, delay, premium dispute, mis-selling, service quality), specific policy and claim references, policyholder contact preference (call back, email, WhatsApp), and the policyholder's preferred resolution. The GRO reference number is read back to the policyholder on the same call. IRDAI mandates GRO resolution within 15 days; the agent sets an expectation of 10 business days for Tier 1 grievances.

For unresolved GRO complaints (beyond 15 days or unsatisfactory resolution): (a) IGMS (Integrated Grievance Management System) / Bima Bharosa — online portal for escalation; IRDAI monitors response times and penalises insurers with high escalation rates; (b) Insurance Ombudsman — jurisdiction up to Rs 50 lakh, free for policyholders, binding on insurers for claims under Rs 30 lakh; (c) Consumer Forum — for disputes involving mis-selling, unfair trade practices, or amounts above Rs 50 lakh.

For health insurance-specific grievances: the agent applies the IRDAI Health Insurance Regulations 2016 protection framework — specifically, Section 20 (prohibition on cancellation except for fraud), non-renewal rights for senior citizens, and the waiting period credit protection on portability. When the grievance involves a repudiation under a waiting period claim, the agent confirms whether the policy was portable (in which case waiting period credit applies) before confirming the repudiation is valid.

Health insurance AI voice deployments operate across two regulatory frameworks simultaneously: TRAI TCCCPR 2018 (telecom regulatory framework for commercial communications) and IRDAI servicing regulations (insurance regulatory framework). Non-compliance with either triggers financial penalties, regulatory action, and reputational risk.

TRAI classification for health insurance: (a) Transactional — calls to policyholders about their active policy (pre-auth confirmation, claims status update, payment due, document request, benefit inquiry callbacks) — DND-exempt, no time restriction beyond the 3-call/day limit; (b) Service Implicit — IRDAI-mandated renewal reminders, claim settlement notices, policy issuance confirmations — DND-exempt and cannot be suppressed even if the policyholder requests it; (c) Promotional — calls to non-policyholders for new policy sales or cross-sell of products not related to their active policy — subject to DND scrubbing, 9 AM–9 PM window, prior consent.

The tricky classification: follow-up calls to lapsed policyholders for revival (after the grace period ends) shift from Transactional to Promotional under TRAI classification. The Kallix platform flags the lapse date and automatically switches the TRAI category for affected policyholders — preventing an inadvertent DND violation that is common in manual operations.

IRDAI servicing records: all policyholder calls must be recorded and available for IRDAI inspection for 90 days minimum. Call recordings must be retrievable by policy number and date. IRDAI's Digital Insurance Regulations 2023 permit digital-only policy records (no physical policy copy required) — the AI agent's call record serves as the consent evidence for digital communications.

For DPDP Act 2023 compliance: health data discussed on servicing calls is sensitive personal data under the Act. The AI agent must operate under the insurer's DPDP privacy notice, which the policyholder accepted at policy inception. No health data is shared with third parties without explicit consent, and policyholders can request data erasure — the agent routes DPDP data requests to the insurer's Data Protection Officer.

Health insurance policy servicing is one of the highest-ROI AI voice deployment categories in the insurance sector — because the call volume is high (active policyholders have frequent servicing needs), the queries are structured and data-retrievable (benefit inquiries, waiting period checks, network lookups), and the cross-sell opportunity (sum insured adequacy, super top-up, critical illness rider) is naturally embedded in the conversation.

Cost savings: human agent handling of a health insurance servicing call costs Rs 320–520 per call (blended: salary, training, infrastructure, supervision, attrition replacement). AI agent cost: Rs 80–140 per call including telephony, LLM inference, and PAS/TPA API calls. At 15,000 servicing calls/month, the monthly saving is Rs 36–57 lakh on an investment of Rs 10–16 lakh, producing payback in 2–4 months.

Complaint reduction: 35–50% reduction in IGMS escalations (production data from Kallix health insurance deployments) from proactive room-rent advisory, pre-hospitalization sub-limit disclosure, and real-time grievance registration with reference numbers. Each IGMS complaint costs the insurer approximately Rs 8,000–12,000 in resolution cost — complaint reduction provides a second revenue stream in the ROI model.

Cross-sell revenue: the sum insured adequacy advisory on servicing calls (not sales calls) achieves 22–31% super top-up attachment rate at scale — because it is delivered in the context of a genuine service interaction, not a sales pitch. At an average super top-up premium of Rs 8,000–12,000 per policy, 10% attachment on 15,000 monthly servicing calls generates Rs 12–18 lakh per month in incremental premium revenue.

NPS improvement: +20–28 NPS points from same-session resolution, proactive sub-limit disclosure, and e-health card delivery in under 5 minutes versus 3–7 day physical delivery — driven primarily by elimination of the runaround between insurer and TPA.

Wellness benefit programs are a rapidly growing differentiator in Indian health insurance — HDFC Ergo WellSurance, Aditya Birla Activ Health Platinum, Niva Bupa ReAssure, and Star Health's wellness program each offer annual health check-ups, fitness tracker integration, step-count rewards, and premium rebates for healthy behaviour. These benefits are consistently under-utilized because policyholders forget they exist.

AI agent wellness inquiry flow: (a) reads wellness program enrollment status from PAS, (b) confirms current wellness credit balance and how points were earned (step count synced from Google Fit/Apple Health, annual health check-up completion, BMI recording, vaccination update), (c) explains redemption options (premium discount at next renewal — typically 5–15% off the renewal premium; additional sum insured credit; or partner healthcare services at no additional cost), (d) identifies unused annual health check-up entitlement and books an appointment with a partner diagnostic centre via API (Thyrocare, SRL, Dr Lal PathLabs).

For corporate group health plans with wellness programs, the agent confirms the employer's annual wellness budget per employee, the enrolled platform (Eka Care, MFine, Maxin wellness), and the remaining balance — plus reminds employees that wellness credits typically expire at the group policy anniversary (not the calendar year end, which is a common confusion).

Wellness inquiry calls have the highest NPS impact of any servicing call type — because policyholders calling to ask a routine question discover value they didn't know they had. The NPS lift from wellness benefit disclosure on routine servicing calls is +8–12 points above baseline in production deployments.

Senior citizen policyholders (65+) represent 18–22% of health insurance inbound servicing volume but account for 45–55% of escalated complaints — primarily because standard IVR and chatbot interfaces fail them. AI voice agents with configurable speech and confirmation protocols are significantly better than IVR for seniors, but must be deliberately tuned.

Conversation adaptations: (a) speech rate reduced to 90–100 WPM, (b) key information repeated twice without prompting, (c) medical jargon translated to plain language (e.g., 'proportionate deduction' becomes 'if you choose a room that costs more than Rs 5,000, the hospital charges you extra on other bills too'), (d) confirmation after each block ('I've told you the room rent limit is Rs 5,000. Is that clear before I continue?'), (e) proactive offer of callback at the senior's preferred time when a process is complex.

IRDAI protections specific to senior citizens: (a) IRDAI Health Insurance Regulations 2016 prohibit health insurers from cancelling or refusing renewal for policyholders above 65 who have been continuously insured — the agent confirms this right when a senior reports receiving a non-renewal notice, (b) co-payment waiver for 65+ with 3+ years continuous coverage (insurer-specific, not universal), (c) premium loading caps on renewal (insurer-specific, governed by product filing).

Digital action alternatives: for actions requiring digital completion (e-health card, digital policy document, online payment, OTP authentication), the agent always offers a branch visit alternative with the nearest branch address queried from the CRM, or the option to add an authorized family member's contact for assisted digital access. This prevents 22–30% of senior citizen drop-offs that occur on digital-only self-service flows.

Chronic condition policyholders represent the highest claim cost segment for health insurers but also the highest retention value — because they have genuine coverage need, high policy loyalty when served well, and significant cross-sell potential for top-up policies. Poorly served chronic policyholders are the most likely to escalate to IGMS and the most likely to port to a competitor insurer.

Waiting period specifics for chronic conditions: if diabetes and hypertension were declared at enrollment, the standard PED waiting period (24–48 months) applies to hospitalizations directly caused by these conditions. However, complications of diabetes (e.g., diabetic retinopathy, nephropathy, neuropathy) are typically treated as separate specific illness waiting periods — the agent reads the product-level waiting period matrix to confirm which complications are covered and from when.

Chronic disease management programs: HDFC Ergo, Niva Bupa, Aditya Birla, and Star Health all offer structured DMP programs for diabetes and hypertension — typically including annual HbA1c, lipid panel, renal function tests, and ophthalmology check-up at partner labs. The agent confirms DMP enrollment, books the annual screening via Thyrocare/SRL API, and routes DMP coordination queries to the insurer's DMP coordinator contact.

Annual health check-up for chronic conditions: most health plans include an annual health check-up in-policy for policyholders above a certain age (typically 45+), covering the standard wellness panel. For diabetic policyholders, the agent confirms whether the policy's annual check-up package includes HbA1c and kidney function tests specifically — which are relevant for diabetic monitoring. If not, the agent recommends the insurer's diabetes add-on rider.

Section 80D advisory calls spike in January–March during the income tax declaration season — employees making IT declarations, self-employed individuals preparing advance tax computations, and salaried individuals submitting investment proofs to their employer. AI agents that handle 80D queries at scale without queue build-up provide significant peak-season capacity relief.

Section 80D limits (as of FY 2024-25): (a) self, spouse, and dependent children: up to Rs 25,000 deduction on health insurance premium; (b) parents below 60: additional Rs 25,000; (c) parents aged 60 or above (senior citizen): additional Rs 50,000; (d) senior citizen without health insurance but with medical expenses: Rs 50,000 deduction on medical expenses (receipts required); (e) total maximum deduction in one FY: Rs 1,00,000 (if policyholder and both parents are senior citizens).

The AI agent's tax advisory flow: (a) reads total premium paid in the FY from the PAS payment ledger, (b) identifies which members are covered under the policy (self/spouse/children vs parents — the premium must be split if both own-family and parental cover are on the same policy), (c) confirms which family members are senior citizens (60+) for the enhanced Rs 50,000 limit, (d) dispatches the tax receipt PDF (Form with insurer letterhead, digital signature, FY-specific premium breakup) to email + WhatsApp.

For group health: the employer-paid premium is not eligible for 80D deduction by the employee. However, if the employee pays for an additional top-up or personal health plan, that premium is eligible. The agent confirms the distinction and confirms whether the employee's personal top-up (if any) is captured in the PAS for tax receipt generation.

Policyholders frequently confuse health insurance and personal accident cover because both involve hospitalization — but the claim pathway, documentation, and cause requirement are fundamentally different. AI agents trained on both product types prevent misrouting that causes claim delays.

Personal Accident (PA) cover: covers hospitalization, temporary total disability (TTD), permanent total disability (PTD), permanent partial disability (PPD), and accidental death — but only for events caused by 'accidental, external, visible, and violent' bodily injury. A heart attack during exercise is not covered by PA (not external/violent); a fall from a ladder is covered. The AI agent confirms the incident type before initiating an FNOL.

PA hospitalization benefit: typically Rs 500–Rs 2,000 per day for the duration of hospitalization (daily cash benefit structure), or a lump-sum benefit linked to the medical expense (indemnity structure). This is separate from the hospitalization coverage under health insurance — a policyholder can claim both if they have both policies. The agent confirms both policy records and initiates claims under both where applicable.

PA add-on riders on health insurance: many health policies include PA riders (accidental death, accidental PTD) that activate for accidental incidents. The agent identifies the active riders from the PAS and initiates the appropriate FNOL — PA rider claim for the accidental death/disability element, health insurance claim for the hospitalization cost.

For motor accident hospitalizations: the agent confirms both the health insurance claim (for hospitalization cost) and the Third Party Motor Insurance claim (for TP bodily injury compensation under Motor Vehicles Act Section 163A or MACT). These are parallel, non-exclusive claims — the health insurer typically subrogates the TP recovery from the motor insurer.