New Way Now: NoBroker leverages Google Cloud AI for broker-free real estate
14Featured in this video: Akhil Gupta, Founder, NoBroker.com
Summary: Akhil Gupta, Founder of NoBroker.com, the world's largest real estate platform without brokers, explains how Google Cloud is enabling them to revolutionize property transactions. NoBroker leverages Google Cloud's AI-optimized infrastructure, including Vertex AI, Gemini models, and L4 GPUs, to automate complex real estate processes and deliver broker-free services. By implementing AI-driven solutions like their Convozen AI and virtual agents, NoBroker automates tasks such as property verification, moving estimations, and customer interactions. This partnership has allowed NoBroker to save customers $1 billion in brokerage fees annually, serve close to 40 million customers, and handle a significant portion of customer service calls with AI-driven virtual agents, ensuring continuous innovation and efficiency without significant infrastructure overhead.
Challenge: Historically, buying or renting property was inextricably linked with brokers and high intermediation costs, requiring extensive "feet on street" effort. NoBroker aimed to overcome this by removing brokers and automating complex, labor-intensive processes like property verification, moving estimations, and multi-language customer service.
Solution: NoBroker built a C2C real estate platform that eliminates brokers. They extensively leverage Google Cloud AI, including Gemini models, to automate tasks such as assessing household items for moving estimations. Their proprietary Convozen AI platform, powered by Gemini and Google Cloud's L4 GPUs, converts diverse customer conversations (email, chat, WhatsApp, voice in multiple Indian languages) into machine-readable formats for analysis and insights. Vertex AI is critical for enabling fast POCs, deploying solutions at scale and speed, and hosting both Google and open-source models without worrying about infrastructure.
Results: NoBroker has successfully broken the norm of brokerage, saving customers $1 billion annually and serving close to 40 million customers. Their AI-driven virtual agents can converse in multiple Indian languages and are projected to handle 25-40% of future call loads, freeing human agents for more complex tasks. They process 10,000 hours of customer recordings daily with near real-time transcription. This robust AI infrastructure enables continuous innovation at a fast pace and ensures cost-effectiveness and speed in model deployment.
Key Insights from Akhil Gupta, Founder, NoBroker.com:
→ "At NoBroker, we have built something called convozen AI. Convozen is all about Zen out of customer conversations. And customer conversations can happen on email. With Gemini, we have automated it fully. We process 10,000 hours of customer recordings on a daily basis. What we are able to do, because we have deployed our models on L4 GPUs.
→ "The way we choose the model is how effective it is in the cost and how fast it can solve the problem."
→ "I think that's where the Google Cloud and the Vertex AI fits in beautifully because, we are able to deploy not only the Google models. We can also host open source models and we can create our own models. So with that, we are able to continuously innovate at a fast pace because we do not have to worry about the deployment and the infrastructure and the scale."
Google Cloud products used: Vertex AI, Gemini, and L4 GPUs
Learn more:
→NoBroker is using AI to match homeowners and home seekers with the end-to-end services they need: https://goo.gle/40tNNZv
→Building a sustainable broker-free real-estate marketplace with Google Maps Platform: https://goo.gle/4lBLhJ6