Sharad Baidya

Decomposed a monolithic app into 15 microservices, heavily relying on Kubernetes. Each handled specific functions like catalog, checkout, orders, product management, and marketing. Customer-facing apps and stakeholders had separate namespaces. Helm charts were created for each service. Istio managed traffic between services. Google Cloud Pub/Sub enabled asynchronous communication. Cloud SQL housed our main DB with multiple replicas. For read-only operations, services accessed replicas. Sidecar proxies provided authorization for each service, enhancing security.

As a solution architect tasked with revitalizing Namshi's catalog, I faced the challenge of accommodating the massive influx of traffic, particularly during peak periods like Black November, when we experienced nearly 10,000 requests per minute. Namshi, a leading eCommerce platform in the Middle East, demanded an agile and cost-effective solution.

The existing catalog infrastructure relied heavily on Cloud SQL, which proved inefficient and costly to scale, especially with the need to increase read replicas. To address this, I spearheaded a comprehensive architectural overhaul. Instead of directly querying Cloud SQL, we introduced a caching layer powered by Redis. Leveraging Redis's in-memory storage capabilities significantly accelerated data retrieval, ensuring swift response times even during peak loads.

Furthermore, utilizing Kubernetes allowed us to scale the number of Redis pods seamlessly, optimizing resource allocation and enhancing system resilience. This strategic shift bolstered performance and resulted in substantial cost savings for Namshi.

Implemented a service that updates the DNS record when config values are updated. Upon changing the service name, we would update the config values and restart the pod. This was implemented using Kubewatch and Go. As a result, there is a seamless connection on the update of the service name.

I designed and developed a versatile questionnaire application tailored for clinical trials, allowing users to create diverse questionnaires with varying answer types. This dynamic platform caters to the extensive needs of clinical trials, accommodating various questionnaire formats and response structures.

Notably, this application has been adopted by leading pharmaceutical giants such as Pfizer, AstraZeneca, and Lilly, underscoring its reliability and efficacy in the industry. The app has played a pivotal role in streamlining data collection processes and driving advancements in medical research by providing a comprehensive solution that addresses the complexities of clinical research.

I spearheaded the development of an auto-build application for Medable, a renowned company specializing in clinical trials and drug validation. This innovative tool revolutionized the process of creating participant schedules by automating each participant's journey.

Traditionally, these schedules involved manually creating tasks and visits, consuming valuable time and resources. Leveraging the power of Node.js and MongoDB, I engineered a robust solution that streamlined the entire process. The auto-build app intelligently generates participant schedules, meticulously defining visits and associated tasks precisely and efficiently.

By automating what was once a labor-intensive task, this solution dramatically reduced the significant amount of manual work from days to mere minutes. This not only optimized resource allocation but also empowered Medable to expedite their clinical trial processes, accelerating the pace of drug validation and medical advancements

I spearheaded the integration of Elasticsearch at Namshi to elevate our search capabilities. By leveraging Elasticsearch, we empowered users with comprehensive full-text search functionality without the need to query the database directly. This not only streamlined the search process but also facilitated keyword-based searches and accommodated variations in spelling.

Implementing Elasticsearch significantly enhanced the overall search experience for Namshi's customers. With the ability to efficiently retrieve relevant results and support various search queries, we ensured a more intuitive and user-friendly search interface, ultimately enriching the browsing and shopping experience on the platform.

ArtOps is an eCommerce for art galleries. As a full-stack developer, I actively set up DevOps pipelines, which played a crucial role in the startup's evolution to a fully functioning entity.

I contributed to every stage of the journey, ensuring seamless growth from ground zero to operational success. Additionally, I focused on optimizing SQL queries, achieving notable reductions in latency and enhancing the overall system efficiency and performance.