magicpin Builds India's Largest Hyperlocal Retail Platform on MongoDB

Despite its trillion-dollar economy, 90% of retail consumption in India still takes place offline. While online retail in India has grown in recent years, much of it still consists of dark stores (a retail outlet or distribution center that exists exclusively for online shopping) and warehouses, the majority of retail establishments—fashion, food, dining, nightlife, and groceries—thrive as physical stores. What’s more, businesses looking to transition to online models are hindered by major platforms that focus primarily on clicks rather than encouraging transactions.

This opportunity was the inspiration for the founders of magicpin, India’s largest hyperlocal retail platform. magicpin has revolutionized the conventional pay-per-click model, where businesses bid on keywords or phrases related to their products or services and then pay a fee each time someone clicks on an ad, with a new pay-per-conversion strategy. In a pay-per-conversion model, businesses only pay when they make an actual sale of a product or item. magicpin does not rely on dark stores, warehouses, or deep discounting; instead, it collaborates with local retailers, augmenting foot traffic and preserving the essence of local economies. This unique model ensures that consumers not only enjoy existing in-store benefits, but also receive additional perks when opting to transact through magicpin.

“We enable the discovery of those merchants,” says Kunal Gupta, senior vice president at magicpin. “Which merchants in your local neighborhood are selling interesting stuff? What’s their inventory? What savings can we offer to buyers? We have data for everything.”

Effectively three SaaS platforms in one, magicpin is a seller app, a buyer app, and a developing logistics app on the Open Network for Digital Commerce (ONDC), which is backed by the Indian government. With over 10 million users on its platform (covering the majority of Indian cities and over 100 localities), magicpin has established itself as a leading offline retail discovery and savings app. magicpin currently has 250,000 merchants in categories ranging from food to fashion to pharmacy.

The power behind magicpin has always been MongoDB's flexibility and scalability. And from the company’s start in 2015, it became clear that magicpin was on to something special.

“In the first week of March 2023 when we onboarded ONDC, we hit almost 10,000 transactions a day. In October last year, we peaked at 50,000 orders in a single day, which is a huge milestone,” says Kunal.

“When an ONDC order is placed, it flows through us. We manage the entire process—from sending the order to the merchant, assigning logistics personnel for pickup and delivery, to handling any customer support tickets that may arise. It's the seamless integration of these elements that defines our contribution to the intricate framework of ONDC."

Having launched using the community version of MongoDB, Kunal realized that magicpin needed to make better use of its relatively lean tech team and allow them to focus more on building the business. He also saw that a managed service would be a more effective way of handling maintenance and related tasks.

“We realized there had to be a better solution. We can’t afford to have all the database expertise tied up with a team that’s focusing on creating businesses and building applications,” said Kunal. “That’s when we started to use MongoDB Atlas."

magicpin uses a multitude of technologies, to store over 600 million SKUs, and handle its SaaS platform, session cache, card, and order management, and MongoDB Atlas sits at the heart of the business.

“For our operational and scaling needs, it’s seamless,” Kunal concludes. “Availability is high, and monitoring and routing are super-good. Our lives have become much easier.”

Watch the full presentation on YouTube to learn more.

← Previous

Technical Services Tools: Embracing Modern Frameworks and Influencing Efficiency

The Technical Services Tools team at MongoDB plays a pivotal role in ensuring efficiency throughout the customer lifecycle. Led by experienced engineers like Jarrod Hinson , the team leverages advanced technologies such as generative AI and modern frameworks like MERN and MEAN to develop tools that enhance security, streamline support processes, and integrate seamlessly into MongoDB's ecosystem. By prioritizing collaboration and innovation, the team not only supports internal stakeholders, but also delivers significant value to MongoDB's customers, setting new standards in technical service excellence. Read on to learn more about working on the team from Jarrod’s perspective. My journey as a Technical Services Tools Engineer at MongoDB Hey there, I’m Jarrod, a Staff Software Engineer on MongoDB’s Technical Services Tools team. Before joining, I worked as a Salesforce engineer, tinkering with custom Full Stack technologies like MERN and MEAN stacks. My background includes stints in healthcare software engineering and service in the US Army. I initially joined MongoDB as a Salesforce engineer in 2017, transitioning to the Technical Services Tools team in 2022 to delve deeper into Full Stack development and engineer applications using gen AI. Influencing tools and security practices In my role as a Staff Software Engineer at MongoDB, I influence internal tools and security best practices across the entire customer lifecycle. Central to my role is ensuring the secure storage of sensitive data and strictly controlling access to resolve customer issues. I craft robust authentication, authorization, and logging systems that track interactions with customer data from both ends, upholding rigorous standards of security and accountability. Embracing modern frameworks and gen AI My work involves extensive use of modern frameworks and gen AI. We deploy the MERN stack with Next.js and MongoDB's Leafygreen UI framework, complemented by Salesforce architecture for our data needs. Leveraging cutting-edge technology with gen AI, we employ MongoDB for databases and vector stores. Customer and internal support is enhanced through our work with assisted gen AI which refines Technical Services case work to boost efficiency and satisfaction. For example, our public Support Portal lets customers chat about any issues they’re experiencing with responses provided through Retrieval-Augmented Generation (RAG) system results pulled from the MongoDB Technical Services Tools database. We utilize RAG for documentation, knowledge base articles, cached entries, and corrective guidance from our custom-built LLM feedback responses. Cross-collaboration opportunities Collaboration thrives at MongoDB. Within our network of customer-facing teams, we work together to define new features and capabilities. For example, we incorporate diagnostic tools into our frameworks through integration efforts with the broader Technical Services team, while partnerships with Cloud and CRM Tech teams manage user permissions across sites. We’ve worked closely with Professional Services leadership to identify existing use cases and seamlessly implement the tools their teams built into our infrastructure. Plus, we’ve helped them identify overlapping customer engagements, quickly understand projects, and enhance customer consultation experience through custom AI summarization. This collaborative approach ensures our tools are robust, secure, and aligned with stakeholder needs, ultimately providing our customers with the best possible experience. Addressing internal stakeholders Engaging with stakeholders across MongoDB presents unique opportunities and challenges. Our projects focus on optimizing processes and enhancing internal capabilities to deliver value to the company and our customers. This role fosters deep cross-departmental relationships, allowing us to solve complex problems with custom tools that drives efficiency across the organization. We provide improved AI triage assistance, case summaries, and next steps to help engineers quickly gain context and engage with customers. We've generated over 100,000 case summaries and have offered triage assistance on over 40,000 cases. Team dynamics Our team is uniquely skilled and operates under a product-oriented delivery (POD) engineering framework broken into five technical areas: Front End, Back End, AI, Salesforce, and Infrastructure. This framework promotes collaboration and efficiency in an effort to drive successful projects. Led by Team Leads and Senior Engineers and guided by our Engineering Manager Lila Brooks , our structure supports continuous learning and engineering best practices. Paul Rooney , our Senior Director, leads by example and takes a hands-on approach with the team's most challenging engineering goals. In my role as a Staff Engineer, I drive multiple projects and aid PODs in achieving their objectives. This structure makes delegation efficient and empowers POD members to make decisions as needed. Why join our team? Our global, primarily remote team embraces flexible work. Yearly in-person offsites encourage team bonding, while the rest of the time, we communicate asynchronously to accommodate global time zones, with essential meetings aligned for overlap. Our leaders prioritize work-life balance and support individual needs to ensure flexibility for all team members. I’ve also found support through MongoDB Veterans, which is an employee resources group that guides veteran applicants to a role within the company aligned to their military experience and brings together a community of supportive co-workers. In terms of day-to-day work on the Technical Services Tools team, what’s most exciting is our specialization in cutting-edge AI applications involving MongoDB processes and data. With rapid expansion, we explore areas like QA, security, AI, Full Stack, and Salesforce, fostering diverse skill sets and knowledge sharing. Projects span intriguing domains from AI applications to custom MERN stack development and Salesforce integration. By joining the team, you’ll have the opportunity to grow your skills, dive into modern frameworks, and join a community of inspiring and talented people! Find out more about Life at MongoDB in technical services by joining our talent community .

July 22, 2024
Next →

AI-Powered Call Centers: A New Era of Customer Service

Customer satisfaction is critical for insurance companies. Studies have shown that companies with superior customer experiences consistently outperform their peers. In fact, McKinsey found that life and property/casualty insurers with superior customer experiences saw a significant 20% and 65% increase in Total Shareholder Return , respectively, over five years. A satisfied customer is a loyal customer. They are 80% more likely to renew their policies, directly contributing to sustainable growth. However, one major challenge faced by many insurance companies is the inefficiency of their call centers. Agents often struggle to quickly locate and deliver accurate information to customers, leading to frustration and dissatisfaction. This article explores how Dataworkz and MongoDB can transform call center operations. By converting call recordings into searchable vectors (numerical representations of data points in a multi-dimensional space), businesses can quickly access relevant information and improve customer service. We'll dig into how the integration of Amazon Transcribe, Cohere, and MongoDB Atlas Vector Search—as well as Dataworkz's RAG-as-a-service platform— is achieving this transformation. From call recordings to vectors: A data-driven approach Customer service interactions are goldmines of valuable insights. By analyzing call recordings, we can identify successful resolution strategies and uncover frequently asked questions. In turn, by making this information—which is often buried in audio files— accessible to agents, they can give customers faster and more accurate assistance. However, the vast volume and unstructured nature of these audio files make it challenging to extract actionable information efficiently. To address this challenge, we propose a pipeline that leverages AI and analytics to transform raw audio recordings into vectors as shown in Figure 1: Storage of raw audio files: Past call recordings are stored in their original audio format Processing of the audio files with AI and analytics services (such as Amazon Transcribe Call Analytics ): speech-to-text conversion, summarization of content, and vectorization Storage of vectors and metadata: The generated vectors and associated metadata (e.g., call timestamps, agent information) are stored in an operational data store Figure 1: Customer service call insight extraction and vectorization flow Once the data is stored in vector format within the operational data store, it becomes accessible for real-time applications. This data can be consumed directly through vector search or integrated into a retrieval-augmented generation (RAG) architecture, a technique that combines the capabilities of large language models (LLMs) with external knowledge sources to generate more accurate and informative outputs. Introducing Dataworkz: Simplifying RAG implementation Building RAG pipelines can be cumbersome and time-consuming for developers who must learn yet another stack of technologies. Especially in this initial phase, where companies want to experiment and move fast, it is essential to leverage tools that allow us to abstract complexity and don’t require deep knowledge of each component in order to experiment with and realize the benefits of RAG quickly. Dataworkz offers a powerful and composable RAG-as-a-service platform that streamlines the process of building RAG applications for enterprises. To operationalize RAG effectively, organizations need to master five key capabilities: ETL for LLMs: Dataworkz connects with diverse data sources and formats, transforming the data to make it ready for consumption by generative AI applications. Indexing: The platform breaks down data into smaller chunks and creates embeddings that capture semantics, storing them in a vector database. Retrieval: Dataworkz ensures the retrieval of accurate information in response to user queries, a critical part of the RAG process. Synthesis: The retrieved information is then used to build the context for a foundational model, generating responses grounded in reality. Monitoring: With many moving parts in the RAG system, Dataworkz provides robust monitoring capabilities essential for production use cases. Dataworkz's intuitive point-and-click interface (as seen in Video 1) simplifies RAG implementation, allowing enterprises to quickly operationalize AI applications. The platform offers flexibility and choice in data connectors, embedding models, vector stores, and language models. Additionally, tools like A/B testing ensure the quality and reliability of generated responses. This combination of ease of use, optionality, and quality assurance is a key tenet of Dataworkz's "RAG as a Service" offering. Diving deeper: System architecture and functionalities Now that we’ve looked at the components of the pre-processing pipeline, let’s explore the proposed real-time system architecture in detail. It comprises the following modules and functions (see Figure 2): Amazon Transcribe , which receives the audio coming from the customer’s phone and converts it into text. Cohere ’s embedding model, served through Amazon Bedrock , vectorizes the text coming from Transcribe. MongoDB Atlas Vector Search receives the query vector and returns a document that contains the most semantically similar FAQ in the database. Figure 2: System architecture and modules Here are a couple of FAQs we used for the demo: Q: “Can you explain the different types of coverage available for my home insurance?” A: “Home insurance typically includes coverage for the structure of your home, your personal belongings, liability protection, and additional living expenses in case you need to temporarily relocate. I can provide more detailed information on each type if you'd like.” Q: “What is the process for adding a new driver to my auto insurance policy?" A: “To add a new driver to your auto insurance policy, I'll need some details about the driver, such as their name, date of birth, and driver's license number. We can add them to your policy over the phone, or you can do it through our online portal.” Note that the question is reported just for reference, and it’s not used for retrieval. The actual question is provided by the user through the voice interface and then matched in real-time with the answers in the database using Vector Search. This information is finally presented to the customer service operator in text form (see Fig. 3). The proposed architecture is simple but very powerful, easy to implement, and effective. Moreover, it can serve as a foundation for more advanced use cases that require complex interactions, such as agentic workflows , and iterative and multi-step processes that combine LLMs and hybrid search to complete sophisticated tasks. Figure 3: App interface, displaying what has been asked by the customer (left) and how the information is presented to the customer service operator (right) This solution not only impacts human operator workflows but can also underpin chatbots and voicebots, enabling them to provide more relevant and contextual customer responses. Building a better future for customer service By seamlessly integrating analytical and operational data streams, insurance companies can significantly enhance both operational efficiency and customer satisfaction. Our system empowers businesses to optimize staffing, accelerate inquiry resolution, and deliver superior customer service through data-driven, real-time insights. To embark on your own customer service transformation, explore our GitHub repository and take advantage of the Dataworkz free tier .

November 27, 2024