Zepto, the fast commerce large, leverages MongoDB Atlas to beat vital scalability challenges. “Behind Zepto’s 10-minute supply promise is a database constructed for velocity!” stated Sachin Chawla, VP for India & South Asia at MongoDB, in a latest social media put up.
By shifting from a monolithic structure with relational databases to MongoDB’s NoSQL platform, Zepto decreased latency for its essential APIs by 40%. This alteration additionally boosted its capacity to deal with six instances extra site visitors, bettering each operational effectivity and the general buyer expertise.
The Problem: Scaling Fast Commerce
Based in 2021, Zepto revolutionised the Indian grocery supply trade with its promise of 10-minute deliveries. Nevertheless, as Zepto expanded—reaching $1.5 billion in annualised gross sales by mid-2024—its legacy infrastructure struggled to maintain tempo.
Efficiency bottlenecks, excessive latency, and an incapacity to ship real-time analytics hindered development and operational effectivity.
A number of months in the past, throughout a presentation at MongoDB.native, Mayank Agarwal, senior architect at Zepto, defined: “We had a giant monolith, and all of the elements have been powered by PostgreSQL and some Redis clusters. As our enterprise grew, we confronted many efficiency points and restrictions on the speed at which we needed to function.”
Zepto’s reliance on Redis clusters for caching additionally grew to become a bottleneck because the enterprise scaled. Kshitij Singh, technical lead at Zepto, famous, “We had a reasonably enormous Redis setup which we needed to do away with, as we weren’t capable of scale it additional. MongoDB’s in-memory caching solved this drawback for us.”
Why MongoDB Atlas?
Zepto sought to interrupt its monolithic structure into microservices and migrate to a NoSQL database to deal with its points. After evaluating a number of databases, MongoDB stood out attributable to its capacity to deal with nested doc buildings and array-based queries successfully.
“The queries have been very performant, given the required indexes we had created, and that gave us confidence,” stated Agarwal. “The most important motivating issue was once we noticed that MongoDB supplies in-memory caching, which may tackle the large Redis cluster that we couldn’t scale additional.”
MongoDB’s Atlas platform additionally provided options equivalent to real-time knowledge archival and analytical nodes, enabling Zepto to separate customer-facing workloads from inside queries. Singh highlighted: “This ensured that buyer efficiency wasn’t compromised.”
Outcomes: Quicker, Scalable, and Dependable
Zepto’s migration to MongoDB Atlas resulted in improved effectivity at an general operational stage. The corporate witnessed an enormous discount in latency, thereby considerably bettering the response instances of essential APIs and enhancing the shopper expertise.
Moreover, the infrastructure scaled to deal with 6x extra site visitors, guaranteeing no efficiency degradation.
The migration to MongoDB additionally led to a 14% enchancment in web page load instances, contributing to larger conversion charges and elevated gross sales. MongoDB’s help for analytical nodes enabled Zepto to keep up constant efficiency, guaranteeing efficient real-time analytics throughout the platform.
The shift additionally streamlined Zepto’s ETL pipelines and archival processes, lowering administration overhead and enabling quicker deployment of latest options.
“Submit the migration, we noticed a latency discount of as much as 40% for key APIs and dealt with throughput spikes, equivalent to throughout campaigns, by enabling knowledge compression and different optimisations. This has made our infrastructure extra sturdy and scalable,” Singh stated.
Scaling New Heights with AWS
Zepto’s success story can be powered by its sturdy partnership with AWS.
“We began Zepto about three years in the past with only one darkish retailer in Mumbai. Immediately, we’ve grown to over 200 darkish shops and annualised gross sales north of INR 10,000 crore,” stated Kaivalya Vohra, co-founder of Zepto. “From day one, we have been clear that AWS was the best way to go.”
AWS’ managed companies and help allowed Zepto to launch new options quickly whereas optimising prices. “AWS helped us obtain a fast MVP with credit from the AWS Activate program. As we scaled, the collaboration with AWS architects enabled us to provide you with options nobody had tried earlier than,” famous Singh.
Zepto leverages AWS throughout the complete buyer journey, from enhancing product discovery to streamlining last-mile logistics. Generative AI options on AWS have additionally enabled quicker decision instances and improved buyer satisfaction.
“AWS has been instrumental in serving to us scale rapidly whereas optimising prices. Its enterprise help and sponsorship for POCs have been invaluable,” Singh added.
The Pricing Dilemma
Zepto is at present underneath fireplace for differential pricing algorithms on Android and iPhones. A number of prospects have expressed their frustration on social media platforms, speculating the doable causes behind the stark distinction in costs.
View this put up on Instagram
A put up shared by Digital Advertising and marketing Advisor | Social Media Coach | Bangalore (@pc_pooja.chhabda)
Some imagine the pricing is influenced by demographic and person machine knowledge, which may probably favour Android customers. The corporate has but to formally touch upon this subject.
The put up How MongoDB Helped Zepto Cut back Latency by 40% appeared first on Analytics India Journal.