IIT Bombay · ARe · 18 years underwriting and actuarial authority in Commercial P&C
Eighteen years underwriting complex commercial risks and pricing them as an actuary — property schedules worth hundreds of millions, engineering projects across emerging markets, reinsurance treaties structured under pressure. The same structural problem repeated itself across every organisation.
The data existed. The pricing models existed. The risk engineering surveys existed. The reinsurance treaties existed. They never talked to each other. The underwriter was the integration layer — manually connecting siloed tools under time pressure at the point of decisions that had real financial consequences.
In 2022, after reaching the ceiling of what traditional insurance roles offered, Dev stepped back and spent four years building systematic Python-based trading systems across equity, commodity, and forex derivative instruments — live capital, real losses, real feedback loops every day.
The discipline of building systems that reason under uncertainty with real financial consequences — and the arrival of AI as a mass-market capability — made one thing clear: the architecture problem in commercial underwriting was now solvable in a way it had never been before.
UnderwriteAI is the result. A platform anchored by The Underwriter's Decision Stack — a proprietary 8-layer framework mapping where AI inserts into the commercial underwriting decision. Ten AI systems being built in public, each one addressing specific layers. Engineering and consulting services grounded in the same framework. All of it published in public so the industry can see exactly how it is done.
A track record of technical precision and regional risk leadership.
Regional Head P&E
Pricing Actuary
CAT Modelling Analyst
B.Tech · Associate in Reinsurance
Practitioner depth meets research rigour.
Dev Vaibhav Markandey is the founder of UnderwriteAI and the author of The Underwriter's Decision Stack framework.
He brings 18 years of operational authority across the full commercial insurance stack — CAT modelling at Flagstone Reinsurance, actuarial pricing at AXA Business Services and RSA Insurance, and commercial property underwriting and reinsurance at AXA Gulf where he held the second-highest underwriting authority in the region after the CEO.
At AXA Gulf he led the regional Property and Engineering book across UAE, Oman, Qatar, and Bahrain — negotiating QS, XoL, and facultative reinsurance facilities, managing spot fac placements as sole regional authority holder, and launching the first parametric insurance product in the GCC as AXA Climate's Regional Champion.
Earlier in his career he built GLM pricing models for AXA UK and AXA Mexico, priced property CAT XoL reinsurance treaties using RMS and AIR at Flagstone Reinsurance, and delivered insurance analytics on assignment for AIG US at Fractal Analytics.
Since 2022 he has built systematic trading systems across equity, commodity, and forex derivatives — developing the engineering discipline to build production systems that reason under uncertainty with real financial consequences. This experience directly informs every AI system built under UnderwriteAI.
He holds a B.Tech from IIT Bombay and the Associate in Reinsurance (ARe) credential.
K. Vibhooti Rajkumar is an Assistant Professor at CSVTU (Chhattisgarh Swami Vivekanand Technical University) and an active research scholar with current research focused on object detection using deep learning.
With over 11 years of experience in Information Technology, her expertise spans data analytics, text mining, image processing, structured data modelling, and parallel processing — capabilities that directly inform UnderwriteAI's approach to extracting structured intelligence from unstructured insurance documents.
At UnderwriteAI she contributes to the AI research foundation of The Underwriter's Decision Stack — supporting signal extraction methodology, model validation approaches, and the application of deep learning techniques to insurance document intelligence problems.
Her work bridges academic rigour and applied AI architecture, bringing research-grade thinking to practical underwriting intelligence systems.
No pitch decks. No sales calls. A direct conversation about your situation. Response within 48 hours.
Engage Now →