Deploys multi-language voice calls through Gemini Enterprise Agent Platform
Cuts voice conversation latency to sub-700ms with Gemini APIs
Processes 100M complex property records inside BigQuery
Auto-fills 75% of mortgage applications through Document AI
Renders massive 3D city digital twins on Compute Engine GPUs
Square Yards uses Google Cloud and Gemini to scale in-house agentic AI, running 200,000 daily calls with sub-700ms latency.
Square Yards uses Google Cloud and Gemini to scale in-house agentic AI, running 200,000 daily calls with sub-700ms latency.
Square Yards outpaces manual dialing constraints
to capture a tenfold lead surge
Buying a home in India is a deeply emotional, sentiment-driven milestone that often requires a cross-family deliberation journey lasting up to three months. As an integrated real estate and mortgage platform, Square Yards coordinates the full property life cycle—from search and transaction to interior design and home financing—for eight million monthly visitors across nine countries.
And because property decisions involve such high stakes, response speed defines competitive survival. If the platform delayed its outreach to an interested buyer by even two days, they risked losing them to a competitor entirely.
So, when customer inquiries suddenly escalated tenfold, the brand hit an immediate operational wall. Outreach relied completely on manual dialing, which structurally capped individual employee output at 150 to 200 connected calls per day.
Compounding this front-end pressure was a severe data bottleneck. Square Yards struggled to ingest and evaluate roughly 100 million legacy property records scattered across fragmented regional registrar offices.
To build a more resilient foundation, Square Yards migrated its core infrastructure to Google Cloud. The company was drawn to the ecosystem's immediate availability of high-performance GPU clusters during a global supply crunch, alongside its unbureaucratic support model. Working directly with dedicated Google Subject Matter Experts (SMEs) to map out solutions, Square Yards bypassed external agencies entirely. Instead, a remarkably compact internal squad of just five to six DevOps engineers successfully managed the entire global infrastructure transition in-house.
Buying a home can be a deeply emotional journey. When inquiries spiked, we used agentic AI not to replace humans, but to intelligently filter for true intent. This ensures our sales executives focus entirely on meaningful, face-to-face closing conversations, driving an expected 1.5x leap in overall business efficiency.
Deepak Kushwaha
Principal Partner and Head of Engineering,
Square Yards
Deploying conversational and agentic layers to maximize human sales efficiency
Square Yards completely modernized its platform blueprint, transitioning legacy workloads into a highly responsive microservices architecture running on Google Kubernetes Engine (GKE) and Cloud Run. To resolve persistent data ingestion limits, the team routed its pipelines into BigQuery. This shift unlocked lightning-fast, automated analytics over their 100 million records, entirely eliminating historical database maintenance loops.
Today, advanced artificial intelligence drives 30% to 35% of the company’s cloud operations. Utilizing the Gemini Enterprise Agent Platform and Gemini APIs alongside Document AI and Vision AI, they built SuperAgent Pro, a proprietary conversational voice engine. SuperAgent Pro automates a complex speech-to-text, sentiment analysis, and text-to-speech pipeline, managing 200,000 automated calls daily across 12+ Indian languages with a sub-700ms response latency. To enable operations at such a large scale, the platform consumes 10 billion Gemini AI tokens every month. When the system pre-screens a serious, high-intent customer, it triggers a live call patch to a human relationship manager, freeing agents to focus strictly on physical site visits.
This agentic architecture extends straight into their mortgage branch, Urban Money. For their mortgage branch, Urban Money, the company is developing a customized WhatsApp bot using Document AI to read unstructured consumer ID card uploads and auto-populate up to 75% of complex, up to 60-field loan documents. For international growth, the company runs compute-intensive procedural modeling on Compute Engine GPUs to combine satellite data and geopositioning, rendering immersive 3D digital twins of entire cities like Abu Dhabi. This secure operational foundation helped the platform achieve ₹1,417 crore in revenue for FY25. As the company looks toward global markets, it aims to list these in-house innovations directly on the Google Cloud Marketplace, solidifying its projected 1.5x efficiency leap into a sustainable engine for long-term growth.
Migrating to GKE and BigQuery eliminated the constant database resizing we faced on legacy clusters. With Google Cloud infrastructure and direct collaboration with Google Subject Matter Experts, our lean DevOps team now runs 200,000 daily automated calls at an ultra-low, sub-700 millisecond conversational response latency.
Deepak Kushwaha
Principal Partner and Head of Engineering,
Square Yards
Square Yards is an integrated real estate and mortgage platform in India. Operating across nine countries, its ecosystem covers property search, transactions, home loans, and interior design.
Industry: Real Estate and Construction
Location: India
Products: Gemini Enterprise Agent Platform, Gemini APIs, Document AI, Vision AI, BigQuery, Google Kubernetes Engine (GKE), Cloud Run, Compute Engine, Secret Manager, Google Cloud Marketplace
