How Databricks Powers Domino’s to Deliver 1.5 Million Pizzas Daily

Over 1.5 million pizzas, delivered across 85 countries every day—just how does Domino’s pull off that feat? The brand was quick to embrace generative AI to streamline its operations and gain valuable customer insights. At the heart of this efficiency lies Databricks, whose Mosaic AI tools power Domino’s latest project, ‘Voice of the Pizza’.

Understanding customer feedback has always been crucial for Domino’s. The Voice of the Pizza project leverages generative AI to analyse feedback from the Domino’s subreddit, providing actionable insights that help improve service quality and product offerings. 

Its main purpose was to assess customer sentiment and identify emerging trends and themes.

“We believe customer feedback is the most reliable source of truth for improving our service,” the Domino’s team noted. By using AI-powered tools, the company can process vast amounts of unstructured data to identify trends, sentiments, and areas for improvement faster than ever before.

The Databricks Tech Stack Behind the Innovation

Bringing tech to life, Databricks’ native products helped Domino’s turn customer feedback into actionable insights. 

Firstly, vector search enabled the rapid retrieval of customer feedback, facilitating quick analysis of sentiment and common themes using semantic search. The team created vector indexes in Databricks to organise and query customer feedback data and then matched the data intelligently. 

Next, the model serving endpoints deployed and served machine learning models for sentiment analysis and feedback classification with optimised performance, enabling autoscaling. 

To help them select the right AI models, Databricks provided Domino’s with its Model Playground. The team could experiment with different AI models, such as Llama, Mistral, and DBRX, fine-tuning them to decide on specific use cases. This helped them prototype a product before deploying it and check out features like function calling and summarisation.

With this infrastructure, Domino’s could process customer insights with unprecedented accuracy, leading to swift and effective decision-making. During this development, the Domino’s team also relied heavily on Databricks Notebooks, which is similar to other AI coders in the market. 

“Additionally, the integration of AI tools such as AI Analyze Sentiment and AI Classify directly into SQL has streamlined our data analysis process, making it easier than ever to extract meaningful insights from our customer feedback,” said a Domino’s team member.

By reducing manual coding and automating tedious tasks, AI allows developers to focus on strategic problem-solving and innovation. The team said that although it is difficult to quantify the exact productivity gains from these tools, they definitely felt them. 

“The in-line code generator transformed our workflow, shifting us from being ‘coders’ to ‘AI directors’,” shared a Domino’s team member. 

Databricks Notebooks played a crucial role in experimenting with different AI models, while AI Playground enabled iterative prompt engineering, further refining the feedback analysis process. These tools collectively sped up development cycles and enhanced efficiency.

By transforming raw customer feedback into structured, actionable insights, Domino’s can refine its services to meet evolving consumer expectations.

Although Voice of the Pizza is not yet in full production, Domino’s sees vast potential in expanding its AI-powered insights across other platforms. To stay ahead of industry trends, the company plans to integrate additional social media channels and competitor feedback analysis.

Not Only Domino’s

As part of its expansion plan in India, Databricks has opened a new office in Bengaluru to capitalise on the country’s rich software industry and abundant pool of world-class engineers. 

Databricks opened its first development centre in Bengaluru in May 2023, contributing to its R&D efforts. Some of the major Indian clients that use Databricks’ AI products are Air India, Krutrim, Aditya Birla Fashion and Retail, Myntra, Meesho, Parle, InMobi, and Freshworks, amongst several others. 

Another interesting one in the Indian context is its partnership with UPL, a multinational company headquartered in Mumbai, which provides sustainable agriculture products and solutions in more than 138 countries.

The company’s UPL Agri-Tech Platform, built in collaboration with Microsoft, uses AI and machine learning to provide crop management, pest control, and nutrient application recommendations. By leveraging Databricks, UPL also aims to enable AI solutions across 20 countries, with the potential to scale up to 50 countries in the future.

The company has also found Databricks’ serverless SQL engine to be efficient, especially in instances where the speed of data pooling is critical. “We have seen it to be quite efficient today because there are a lot of instances where the speed of the data pool is very critical for us,” Vijay Balakrishnan, chief data & analytics officer at UPL Group, told AIM.