Financial services firms accelerating AI adoption across software delivery are seeing early gains in testing and planning, but risk missing the bigger opportunity if they focus only on incremental efficiencies, according to Srini Chelian, VP & Head of Technology, Personal Lines at Definity.
Speaking on the QA Financial podcast ahead of his session at the QA Financial Forum Toronto 2026, which took place April 23, Chelian outlined how the Canadian insurer is embedding AI across its software development lifecycle (SDLC), while warning that true transformation lies in rethinking operating models rather than simply automating existing processes.
Chelian presented “Transforming Delivery and Service: How Definity Is Using AI to Build Better, Faster and Cheaper” on April 23 as part of Track 1 at the QA Financial Forum in Toronto.
Scaling platforms
Definity’s approach to technology transformation has been shaped by rapid growth, including its acquisition of Travelers’ Canadian P&C operations.
Chelian said the organisation has prioritised “scaling platforms” to ensure technology costs do not rise in line with business expansion.
“So one of the things that we’ve always thought of… is we invest in what we call scaling platforms,” he said. “Your technology or the operations cost cannot grow linearly with the business growth because then you’re not actually running a profitable business.”
Cloud adoption, particularly through platforms such as Guidewire and Google Cloud, has enabled the firm to absorb large-scale change without significant incremental cost.
At the same time, Chelian described a shift towards a more self-service model for business teams.
“We don’t want to be an ala carte kind of a business… We want to be more like a buffet where business can actually pick and choose and do more self-service work,” he expained, reserving deeper technology partnerships for more transformational initiatives.
AI delivering early gains
While AI is being applied across the SDLC, Chelian pointed to testing and planning as the areas delivering the most immediate value.
“If I have to answer pointedly, I think the testing phase and the planning phase, we’ve seen greater benefits,” he said, adding that development remains “probably early stages.”
In planning, AI is already being used to process meeting transcripts and design documents to generate requirements, reducing manual effort. In testing, gains are more pronounced, particularly in automation and execution.
However, Chelian cautioned against viewing isolated improvements as true transformation.
“That in itself does not guarantee the business value delivery… you’re still in that journey where you can then translate this to reimagine that SDLC cycle,” he said, emphasising the need to measure outcomes “from concept to cash.”
Chelian described quality engineering (QE) as a “low hanging fruit” for AI adoption, given the structured nature of testing and the availability of historical data.
“We’ve actually used AI significantly in QE… purely because of the structured nature of the function,” he said.
By leveraging repositories such as Confluence and Jira, Definity is using AI to generate test cases, enrich regression testing and build execution plans based on accumulated organisational knowledge.
“Now you don’t have to rely on word of mouth… it’s able to leverage all of those historic information and do that in record time,” he said.
The result is a significant increase in testing throughput without compromising quality.
“It significantly reduces our execution times… which means our release frequency is more and the quality is more,” Chelian added. “So you’re not compromising on quality, but at the same time you’re able to amplify the speed to market.”
Making data ‘AI-ready’
A critical enabler of this shift has been Definity’s data strategy, particularly its move to the cloud and a more targeted approach to data curation.
Chelian warned against overly ambitious, multi-year data lake initiatives, noting that many have failed to deliver expected outcomes. Instead, Definity focuses on use-case-driven data preparation.
“We’re not looking to solve world hunger… you’ve got critical use cases… but you create knowledge bases for the AI models… for those use cases,” he noted.
Key priorities include data accessibility, lineage and explainability, particularly important in regulated sectors such as insurance.
“The audit and governance is big, especially when you have non-humans working on these things,” Chelian added.
“Your technology costs cannot grow linearly with business growth because then you’re not actually running a profitable business.”
– Srini Chelian
As firms begin experimenting with agentic AI, Chelian stressed the importance of distinguishing it from traditional RPA and maintaining human oversight.
“For me, the key difference… is if there’s a learning system within an RPA solution, then it’s getting closer to the agentic AI,” he said.
Definity is already deploying AI-driven capabilities in contact centres, including agent assist, call summarisation and sentiment analysis, while early work is underway in claims.
However, Chelian emphasised that human-in-the-loop models remain essential in regulated environments.
“Because insurance is a regulatory industry and advice is a big part of that, there has to be human in the loop,” he stressed.
Beyond efficiency
For QA and engineering leaders, Chelian’s core message is to look beyond incremental automation gains and focus on structural change.
“I think we need to fundamentally think about how does this change the operating model for testing,” he stated.
This includes preparing for hybrid workforces combining human testers and AI agents.
“Is the future a hybrid workforce where you’ve got agents testing and you’ve got humans testing?” Chelian asked. “We absolutely need to think about a hybrid workforce.”
He warned that focusing solely on efficiency risks missing the broader transformation opportunity.
“We’ll be doing ourselves a disservice by trying to only focus on the incremental benefit,” Chelian said.
Chelian expanded on these themes at the QA Financial Forum Toronto 2026, where his session explored how Definity is applying AI across delivery, testing and customer experience.
As Chelian’s comments suggest, the real challenge for financial institutions is no longer whether to adopt AI in testing, but how to translate early gains into a fundamentally reimagined SDLC, one that balances speed, cost and resilience in an increasingly complex, AI-driven landscape.
WATCH OUR PODCAST WITH SRINI CHELIAN
TOMORROW
NEXT MONTH
Why not become a QA Financial subscriber?
It’s entirely FREE
* Receive our weekly newsletter every Wednesday * Get priority invitations to our Forum events *
REGULATION & COMPLIANCE
Looking for more news on regulations and compliance requirements driving developments in software quality engineering at financial firms? Visit our dedicated Regulation & Compliance page here.
READ MORE
- Inside banking’s shift to smarter QA to tackle complexity and risk
- SmartBear CPTO on AI in banking QA: ‘Impressive metrics but no critical scenarios’
- Banks push beyond traditional QA as resilience testing gains ground
- Banking QA professionals warn AI still doesn’t know ‘where the bodies are buried’
- RECAP: The QA Financial Healthcare & Insurance Forum Philadelphia 2026
WATCH NOW
QA FINANCIAL PODCASTS
CLICK HERE TO LISTEN TO OUR EXCLUSIVE CONVERSATIONS
