Quant AI's Contact Center Agent Cuts Call Times by 26% in Production Deployment

Three minutes doesn't sound like much. But when you're running a contact center handling millions of insurance calls, cutting average call time from 11 minutes 30 seconds to eight minutes 30 seconds is the difference between a cost center and a competitive advantage.

That's exactly what Quant AI's agent Ava delivered at Fortitude Re, a major reinsurance firm. The results were presented at IBM Think 2026 by Chetan Dube, CEO of Quant AI, alongside IBM's Yogendra Goyal. The numbers: 84% of calls resolved without a human agent, first-call resolution jumping from 71% to 86%, and that three-minute drop in average handle time.

The significance isn't just the metrics. It's that these are production numbers, not lab results.

From Pilot to Production: Why This Announcement Is Different

Why Conversational AI Is Getting Real Investment in Enterprise Contact Centers

Financial services and insurance leaders are moving beyond chatbots to deploy agent systems that handle millions of customer interactions.

Mission Media

Most AI announcements in the contact center space come with vendor-estimated projections. This one doesn't. Fortitude Re already had a US$450 million, 10-year partnership with IBM to manage over four million life and annuity policies. Ava was deployed into that existing, live infrastructure.

That matters because it removes the "but will it work at scale?" objection. The policyholder contact center was already operating at scale. Ava stepped into a real production environment and delivered measurable gains.

Industry benchmarks for AI voice agents in 2026 put autonomous call resolution at 60% to 80%. Ava's 84% puts it at or above the top end of that range. The 15-percentage-point jump in first-call resolution is also meaningful: research shows every one-point improvement in first-call resolution saves a contact center roughly US$286,000 annually. A 15-point gain implies something in the range of US$4.3 million in annual savings.

What Ava Actually Does

Ava handles the full range of policyholder support: answering questions about policies, processing payments, sending out forms, and authenticating customers. When a call is too complex for automation, Ava hands it off to a human agent with full context already loaded. No customer retells their story. No dropped context.

It operates in English and Spanish, which is a practical detail that signals how the underlying architecture is built. Multilingual support isn't a cosmetic feature. It reflects the kind of real-world deployment complexity that separates a demo from a production system.

As Dube put it at IBM Think: "Most businesses are missing the connective tissue, the reasoning layer that links systems, people and decisions together. That's what makes our work with IBM different: we're building a true active reasoning conveyor belt that seamlessly connects both AI and humans to unlock real enterprise gains."

IBM's Goyal, who has spent 25 years working with insurance carriers, framed the shift plainly: "It was always about incremental solutions. With Quant, it moves to exponential."

APAC Financial Services Leaders Are Next

IBM Think 2026: Where the Generative AI Enterprise Focus Really Lands

The keynotes and partnership announcements reveal which industries are moving from experimentation to production deployment.

Mission Media

The Fortitude Re case is an insurance deployment, but the underlying pattern applies across financial services, banking, and any business running high-volume customer support.

KPMG data shows 58% of Asia Pacific companies are already using AI in at least limited ways, with that figure expected to hit 80% within two years. IDC describes the region as moving from AI experimentation to industrialization. The Quant-IBM-Fortitude Re deployment is a benchmark for what that industrialization looks like in practice.

Gartner projects conversational AI will save US$80 billion in contact-center labor costs globally by the end of 2026, with an average return of US$3.50 for every US$1 invested. The organizations leading that curve are moving now, not waiting for the market to mature.

The Ava deployment also signals something important about how enterprise AI partnerships are evolving. This wasn't a startup selling a standalone tool. It was a vertical application built on IBM's existing infrastructure, deployed into a live production environment through a pre-existing client relationship. That model, where AI capabilities are layered into established enterprise contracts, is likely how most large-scale deployments will happen in financial services.

For executives evaluating AI contact center investments in Asia Pacific, the most useful question isn't "does this technology work?" The Fortitude Re numbers answer that. The better question is: what production-validated partnerships already exist in your sector, and how quickly can you move from proof of concept to live deployment?