Category archive
Sarvada research, workflows, and commentary focused on ai & insurtech.
Archive
106 published articles
Topic
AI & Insurtech
Latest
August 12, 2026
Browse by topic
106 articles
AI & Insurtech - August 12, 2026 - 10 min read
Plum's August 2026 AI risk suite spans cyber, professional indemnity, D&O, CGL and asset protection rather than offering one AI policy. This is the gap map behind that structure: four AI failure modes traced across the towers an Indian corporate already holds, with the questions to ask each insurer at renewal.
By Sarvada Editorial Team
AI & Insurtech - August 11, 2026 - 10 min read
EarthRe Insurance IFSC Limited, the reinsurance subsidiary of insurtech InRisk Labs, says it is the first incorporated reinsurer licensed by IFSCA at GIFT City. This post examines what a data-native, parametric-first reinsurer adds to an Indian nat-cat programme, and how brokers should test a model-driven quote before the 2027 renewal.
By Sarvada Editorial Team
AI & Insurtech - July 24, 2026 - 11 min read
An advisor has a rich twenty-minute conversation and writes three lines afterward, losing the family details, the policy held elsewhere and the follow-up promised. How recording, transcription and an LLM can turn the conversation into a structured client-book record, with the consent hygiene it demands.
By Sarvada Editorial Team
AI & Insurtech - July 23, 2026 - 10 min read
A vendor's 99 percent accuracy is a single averaged number on the vendor's own clean documents, and it tells a broking firm almost nothing about its own inbox. How to build a ground-truth test from your own documents, measure per field, and write a contract that holds.
By Sarvada Editorial Team
AI & Insurtech - July 21, 2026 - 11 min read
An agent that reads the renewal calendar, assembles the expiring-policy data, drafts the insurer and client chases and escalates non-responses can compress the renewal grind, but only inside hard guardrails: no autonomous binding, human sign-off on anything price-bearing, and an immutable log.
By Sarvada Editorial Team
AI & Insurtech - July 19, 2026 - 11 min read
A broking firm's service mailbox mixes endorsement requests, claim intimations, quote requests and account queries in one queue, and the claim buried in it is a notice-condition failure waiting to happen. How LLM classification turns that queue into routed, timed work.
By Sarvada Editorial Team
AI & Insurtech - July 17, 2026 - 10 min read
A broking firm with thirty insurer agreements accumulated over fifteen years cannot tell you what it is owed without opening files. Contract NLP can read the rate schedules, accrual triggers and set-off rights into a schema, but the arithmetic downstream must stay deterministic.
By Sarvada Editorial Team
AI & Insurtech - July 16, 2026 - 10 min read
Your commission on a policy is not a forecast. It is a calculation with a contracted rate, a defined base and a known set of adjustments, which means it has one right answer. Any tool that hands an advisor an approximation for a number that is exactly computable has quietly decided your money is not worth the arithmetic.
By Sarvada Editorial Team
AI & Insurtech - July 16, 2026 - 10 min read
Insurer commission statements arrive in a dozen formats with no shared key, which is why matching them against your placement register resists automation. An engineering view of parsing, cardinality, confidence bands, exception queues and idempotent re-runs.
By Sarvada Editorial Team
AI & Insurtech - July 15, 2026 - 10 min read
Every insurer prints its schedule differently, half of them are scans, and the tables are not tables. What a machine can actually pull out of an Indian policy PDF, what it quietly gets wrong, and why a confident wrong expiry date is worse for an advisor than no extraction at all.
By Sarvada Editorial Team
AI & Insurtech - July 14, 2026 - 10 min read
Nobody decided WhatsApp would be the surface Indian insurance distribution runs on. It won on install base, zero training, handset tolerance and the ability to carry a PDF. Why the alternatives lost, what that means for tooling, and the four things a chat thread will never be.
By Sarvada Editorial Team
AI & Insurtech - July 14, 2026 - 10 min read
An honest inventory of what one advisor actually runs on: WhatsApp, a phone gallery of policy PDFs, four insurer portals with four logins, one spreadsheet, and whatever app the principal handed over. Which layers hold, which have no owner, and why the choice is rarely the advisor's to make.
By Sarvada Editorial Team
AI & Insurtech - July 13, 2026 - 10 min read
The advisor works in one language, the policy wording is in English, and the client needs a third register entirely. That gap is a distribution problem, not a user-interface preference, and the advisor standing in it is carrying a translation obligation nobody has acknowledged.
By Sarvada Editorial Team
AI & Insurtech - July 12, 2026 - 10 min read
A renewal reminder to a client who already holds the policy is servicing, and servicing is the one outbound act an advisor performs on their own authority. A broadcast offer is not, and a discount is a criminal offence. What is left is four moments, one channel ranking and three sentences.
By Sarvada Editorial Team
AI & Insurtech - July 11, 2026 - 11 min read
Most Indian advisors run their book in a spreadsheet, and that is not a stupid decision. It is a decision with six specific failure modes, three of which are survivable at any size and three of which compound quietly until the loss is already booked.
By Sarvada Editorial Team
AI & Insurtech - July 9, 2026 - 9 min read
How brokers use AI and computer-vision valuation, benchmarked against live construction and plant-and-machinery cost data, to detect underinsured sums insured before renewal so the average clause never bites at claim time.
By Sarvada Editorial Team
AI & Insurtech - July 8, 2026 - 9 min read
ISO's generative-AI exclusion endorsements went live on 1 January 2026 and carriers are fragmenting AI cover across lines. This is a clause-level map of where a hallucination, model drift or automated decision now falls between GL, D&O, cyber and Tech E&O for Indian buyers, and how to negotiate carve-backs at renewal.
By Sarvada Editorial Team
AI & Insurtech - July 6, 2026 - 8 min read
How Indian general insurers can jointly train fraud and network-detection models without moving policyholder records across firms, satisfying the DPDP Act while catching the cross-insurer rings that the IRDAI fraud framework of April 2026 targets.
By Sarvada Editorial Team
AI & Insurtech - July 5, 2026 - 10 min read
How AI buyer-risk models set and revise per-buyer credit limits and forecast receivables default on Indian export and domestic-supply cover, and what brokers must check in the wording before they place it in 2026.
By Sarvada Editorial Team
AI & Insurtech - July 4, 2026 - 8 min read
Machine-learning default-probability scoring on contractor financials, order books and track record is starting to unlock surety capacity in India and speed bid, performance and advance-payment bond decisions after the 2026 framework changes.
By Sarvada Editorial Team
AI & Insurtech - July 3, 2026 - 8 min read
A standalone AI-cover market went live in 2026, pricing model-failure insurance only on documented governance evidence. This is a buyer-and-broker guide to what affirmative AI liability policies actually pay for and the qualification bar Indian GCCs and AI-first firms must clear.
By Sarvada Editorial Team
AI & Insurtech - June 26, 2026 - 6 min read
From FY2026-27 the BRSR Core set of about thirty assured KPIs covers all top-1000 listed companies, a structured ESG dataset that D&O underwriters have barely mined. This post shows how insurers can use AI to pull greenwashing and misstatement signals from BRSR filings, why those signals are a defined D&O trigger, and where the assured-data boundary now sits.
By Sarvada Editorial Team
AI & Insurtech - June 26, 2026 - 6 min read
Indian cyber insurers are replacing static questionnaires with continuous, AI-driven external attack-surface scans and security ratings run before they bind cover. This post explains to brokers how those scans work, how to read and contest a poor rating, where SME maturity gaps trigger declines, and how to prepare a client's external posture ahead of a cyber placement.
By Sarvada Editorial Team
AI & Insurtech - June 26, 2026 - 5 min read
Commercial insurance runs on multilingual paper that English-trained extraction pipelines mishandle: vernacular invoices, FIRs, panchnamas and ledgers. This post shows how the government Bhashini stack and AI4Bharat's IndicTrans2 can be wired into claims and underwriting document processing, the accuracy and script limits to plan for, and the data-residency advantage these public models hold over foreign OCR and translation APIs.
By Sarvada Editorial Team
Sarvada Intelligence
See how Sarvada turns research, risk signals, and operational context into faster commercial underwriting decisions.