India's 4 million developers might find their data platform in MongoDB

In an insightful conversation, Sachin Chawla, Vice President for India and ASEAN at MongoDB, offered a comprehensive look into the burgeoning data platform landscape across India and Asia. He detailed the key trends driving adoption, MongoDB's strategic role in supporting businesses from agile startups to sprawling enterprises, and the company's unwavering commitment to nurturing developer talent in the region.

Our discussion delved into the evolving demands of modern applications, the rapid surge in AI adoption, and the critical factors of scalability, resilience, and security that define contemporary data platforms.

The Evolving Landscape: Apps, AI, and Developer Focus

The conversation kicked off with an overview of the current data platform scene in India and Asia.

"India, as you can see, is home to more than 4 million developers, and we see both India and Asia as regions for high growth," Sachin began. He highlighted several driving forces. "Firstly, customers are building a lot of modern applications and simultaneously modernising legacy applications. Secondly, there's a significant impetus on AI. More and more customers are building AI-infused applications and experimenting with different AI approaches. While it is the start of this revolution, most customers are currently doing basic AI applications, but we also have those building more sophisticated applications for things like customer recommendations."

He also emphasised the growing pressure on software development. "Thirdly, there's a strong focus on developer productivity. The time given to developers to develop and release applications is shrinking, so they have to do more with less. These are common trends we observe across both markets."

MongoDB's customer base in the region reflects this diversity. "Our customer base is broad," Sachin explained. "In India, it ranges from small, single developers to early-stage startups like Ubuys and RFPIOs, to large-scale startups like Zepto, which is one of our largest clients and a public case study. Zepto's order tracking, for instance, is entirely built on MongoDB. We also work with ISVs (Independent Software Vendors) like Intellect AI, a large FSI ISV that has built its multi-agent AI platform, Purple Fabric, on MongoDB to automate and augment operations, risks, and compliance. Then there are large enterprises like banks and Tata Neu, whose application is also built on MongoDB."

Nurturing the Startup Ecosystem: Beyond Technology

India's vibrant startup scene is a key focus for MongoDB, which offers more than just its core technology.

"We do a lot to help startups and their ecosystem," Sachin affirmed. "Beyond the technology itself, we focus on how they can adopt and grow with it. We have consultants and advisors who work day in, day out with them on architecture design. We also drive a lot of 'developer groundswell.' For example, we organise developer days across different cities, offering hands-on labs and design reviews."

MongoDB is also heavily invested in skill development. "We have set a goal to upskill half a million students in universities on MongoDB to build a strong developer base, and currently, over 200,000 students have already taken MongoDB courses," he shared. "We also host weekly training sessions, two to three-day programs where we invite customers and their developers to our offices for training and certification. Additionally, we have a MongoDB community with champions from various organisations. There's a significant effort to develop and upgrade skills."

Common Patterns and the Power of the MongoDB Platform

When asked about the types of applications emerging from Indian startups and which MongoDB platforms resonate most, Sachin highlighted the database's versatility.

"We are a general-purpose database, so you can use us for various use cases," he explained. "You can put any kind of data on us – structured and unstructured data, including video, attachments, and audio. Being general-purpose means we can handle any workload type: transactional data, geospatial data (like tracking a food delivery rider), or time-series data for IoT devices."

He gave specific examples: "Customers use us for diverse applications, from IoT to AI use cases. For example, RFP IO uses our vector search and AI capabilities to help customers respond to RFPs more efficiently by automatically categorizing questions related to security or performance. So, our applications range from payments, order management, and e-commerce order tracking to AI solutions and recommendation engines, and even building operational data layers. It's very broad."

The evolution of the MongoDB platform itself is a key differentiator. "On the platform side, we have evolved significantly," Sachin noted. "The database is the core, where all data is stored. For modern applications, you need Google-like search. Instead of plumbing another search engine like Elasticsearch, we offer Atlas Search, which is native to the platform. For AI, you need vector search to vectorize data... With us, Vector Search is embedded natively."

He further detailed the integration benefits: "Thirdly, customers need embedding models to train their models with their own data. We acquired Voyage AI, which provides these embedding models. The platform has become very important because it integrates all these components natively, eliminating the need for customers to manage multiple disparate systems. Customers struggle with managing one system, let alone five! So, we see them starting with one database, then quickly adopting full-text search, vector search, and embedding models, as well as Atlas 3, which we launched last year."

Voyage AI: A Strategic Acquisition for Trustworthy AI

The acquisition of Voyage AI has been pivotal in enhancing MongoDB's offerings for AI-driven applications.

"When using Large Language Models (LLMs), they are probabilistic," Sachin explained. "If a company, especially in healthcare or financial services, asks a query, they need a more accurate output. This requires training the LLM with their internal data, which is where an embedding model comes in. Voyage AI is that embedding model, now natively integrated into our platform. Voyage AI is arguably one of the best embedding models globally."

He highlighted its crucial benefits: "It helps in two crucial ways: A) it assists in training the model, and B) it significantly reduces hallucinations, which are a common issue with AI models. It also helps with re-ranking results to ensure the most relevant output. All of this is now native within our platform. Customers tell us this truly solves the 'plumbing' problem and provides an embedding model that reduces hallucinations and facilitates re-ranking. Furthermore, Voyage AI comes with pre-packaged embedding models for specific industries like FSI or healthcare, which is another significant benefit."

Enterprise Demands: Scalability, Resilience, and Security

Indian enterprises prioritize specific requirements when adopting a data platform.

"Of course, cost and scalability are crucial, but I'd add two more factors: resiliency and security," Sachin asserted. "There's a lot of discussion around providing a resilient architecture and ensuring security. Another key aspect is performance, specifically performance at scale. The scale we talk about in India is on a different level."

He provided a compelling example: "Take Zepto as an example. They were previously on a monolithic architecture using SQL. When they moved their application from SQL to MongoDB, the latency dropped by 40%, and they could handle six times more traffic with that reduced latency. So, performance at scale is extremely important."

Regarding resilience, Sachin elaborated, "We provide a three-node architecture (active-passive-passive). If the active node goes down, a passive node automatically takes over. You can deploy these across three different availability zones within AWS or other cloud providers. We offer the platform across all three major cloud providers (AWS, GCP, Microsoft Azure), and you can even have instances of the same application in two different clouds, providing immense residency and flexibility."

Upskilling the Next Generation: Partnerships in Academia

MongoDB's commitment to developer upskilling extends significantly into academia.

"There are two main pieces to this," Sachin explained. "Firstly, we have MongoDB University, where anyone can go and take a course today. Secondly, we have partnered with AICTE (All India Council for Technical Education) and various universities across India. Our collaboration with universities involves not just training students but also a 'train the trainer' approach, where we train professors. We've also partnered with GeeksforGeeks. So, there are three or four initiatives working in tandem to reach and train these students."