For decades, housing finance in India remained fundamentally paper-driven. That model is now changing rapidly.
The next phase of mortgage lending is being shaped not merely by branch expansion or cheaper capital, but by AI-led underwriting, predictive analytics, automation, and digitally orchestrated customer journeys. Mortgage companies are steadily evolving into technology-led financial platforms where data intelligence is becoming as critical as lending itself.
At the centre of this transition is Anubhav Rajput, CIO & CTO, PNB Housing Finance, who is leading a large-scale technology modernisation initiative at the organisation. From cloud-native infrastructure and AI-enabled collections to predictive underwriting and mobile-first onboarding, the company is reimagining what a modern mortgage institution could look like in India’s rapidly evolving housing market.
The next housing finance boom will be built beyond metros
According to Rajput, the strongest momentum in housing finance is now emerging from Tier-II, Tier-III, and even Tier-IV cities, where mortgage penetration remains low, but demand growth is accelerating rapidly.
“The affordable housing segment is seeing tremendous expansion because penetration in those markets is still very low while growth rates remain significantly higher,” he says.
At the same time, another trend is reshaping urban housing finance: the rise of hyper-luxury real estate. Customers purchasing premium properties increasingly expect highly personalised and digitally seamless experiences across the buying lifecycle.
This duality is forcing mortgage companies to manage two very different realities simultaneously: high-volume affordable lending, where automation and scalability are critical; and premium lending, where customer experience becomes equally important.
But perhaps the biggest shift underway is digitisation itself.
Processes such as legal title searches, technical valuations, customer verification, and tele-personal discussions are steadily becoming digital. According to Rajput, processes that were around 50–60% digitised two years ago are now moving closer to 85–90% digitisation.
“The mortgage journey itself is steadily transforming into a digitally orchestrated workflow,” he says.
Why affordable housing needs a different technology architecture
For lenders, the challenge is no longer simply digitising processes. The real challenge is scaling underwriting intelligence across fragmented borrower profiles.
A salaried metro borrower behaves very differently from a self-employed tailor, driver, or small-business owner in an emerging market. This becomes especially important in affordable housing, where loan sizes are smaller but underwriting risks are often more complex.
To address this, PNB Housing Finance has built specialised rule engines capable of evaluating customers at a profession-specific level.
“If the applicant is a driver, the system evaluates ride history, frequency, and customer ratings. For a tailor, it assesses stitching volumes, pricing patterns, and earning potential before calculating repayment capability,” Rajput explains.
The company has also digitised onboarding through Aadhaar verification, DigiLocker integration, PAN authentication, E-NACH mandates, and mobile-based onboarding workflows.
Its sales workforce now operates through a mobile-integrated platform connected to fintech ecosystems and account aggregators, allowing the lender to expand into underserved markets without proportionally increasing operational complexity.
Mortgage collections are becoming predictive
One of the biggest shifts underway in housing finance is the transition from reactive collections to predictive collections.
Traditionally, lenders responded after delinquency occurred. AI is now allowing mortgage companies to anticipate repayment stress before defaults happen.
At PNB Housing Finance, predictive models continuously analyse repayment behavior and identify the probability of default at an individual customer level.
Based on these signals, some customers receive digital nudges via WhatsApp or phone; high-risk accounts may be escalated directly to field teams, while low-risk customers may require minimal intervention.
“We are increasingly using models that can predict the propensity of a customer to default in a given month,” Rajput says.
This becomes especially important in affordable housing, where operational efficiency directly impacts profitability.
AI moves from experimentation to measurable ROI
Rajput believes the industry is now moving beyond AI experimentation toward measurable business outcomes.
“We do not deploy AI for the sake of AI. Every implementation must demonstrate measurable ROI,” he says.
One of the company’s most successful AI initiatives has emerged in the sanctioned-but-undisbursed loan category. In housing finance, customers often secure sanctions but delay disbursement due to documentation gaps, property-related dependencies, or better competing offers.
To solve the problem, the company deployed an AI-led outreach engine capable of proactively engaging customers, identifying bottlenecks, and recommending corrective actions dynamically.
The results were significant: nearly 32–33% of sanctioned-but-undisbursed loans converted into actual disbursements; the AI engine identified customers receiving better rates elsewhere; and the system dynamically triggered retention interventions such as revised pricing or enhanced eligibility evaluations.
“These insights help us optimise customer engagement and improve our ability to convert sanctioned loans into monetised disbursements,” Rajput says.
AI is also reshaping customer service. Today, nearly 18–19% of customer service interactions at PNB Housing Finance are resolved automatically through AI-enabled systems without requiring human intervention.
Routine requests such as statement downloads, foreclosure support, address changes, and nominee updates are increasingly handled through NLP-powered voice and digital interfaces.
“The distinction between a human agent and an AI voice agent will increasingly blur,” Rajput says.
Modernising while running at full speed
According to Rajput, one of the biggest challenges was replacing legacy systems without disrupting ongoing operations.
“You are effectively rebuilding the engine while the vehicle is still moving at 60 miles per hour,” he says.
Over the past three years, PNB Housing Finance has modernised most of its core platforms, migrated nearly 85–86% of workloads to the cloud, built strong in-house engineering capabilities, and adopted modern DevOps practices.
Today, the organisation executes nearly 9–10 production releases every month—far higher than traditional release cycles in many financial institutions.
Its sales platform, currently used by nearly 6,000 employees, has been developed entirely in-house, allowing faster enhancement cycles and significantly lower development costs.
The company has also invested heavily in operational resilience. Rajput points to a recent example where the organisation operated entirely from its disaster recovery environment for nearly 12 days without impacting users or branch operations.
The next phase of AI in housing finance
While underwriting, collections, and servicing remain the most visible AI use cases today, Rajput believes the next wave of transformation will emerge in areas such as fraud analytics, property intelligence, geographic expansion modeling, and operational forecasting.
One particularly interesting use case under evaluation involves AI-powered drone analytics for plot-plus-construction loans. The proposed solution combines drone-based inspections, historical imagery, video analytics, and AI-powered image comparison to determine whether construction milestones have been completed.
“It allows us to improve compliance, reduce fraud risk, and significantly improve operational efficiency,” Rajput says.
The company is also exploring AI-driven expansion models capable of evaluating demographic patterns, competitor presence, delinquency trends, and regional economic signals before entering new markets.
As India’s housing finance ecosystem evolves, Rajput believes the competitive battleground is steadily shifting away from branch density and toward digital intelligence.
For India’s mortgage industry, the transformation is no longer about digitising paperwork. It is about rebuilding housing finance itself around AI, automation, and intelligent platforms—and that transition is only beginning.