I’m Shubham Soni — a hands-on software engineer and technical leader building cloud-native platforms, scalable APIs, and full-stack products. I work across architecture and implementation to make systems faster, more resilient, and easier for teams to evolve.
I like solving hard technical problems with clear architecture, pragmatic trade-offs, and measurable results.
“Build for today. Design for change. Make reliability part of the product.”
A principle I bring to architecture and delivery.
Over the course of my career, I’ve worked across SaaS and enterprise products — from frontend integrations and API contracts to event-driven backends, cloud infrastructure, and production reliability.
I enjoy simplifying complex systems, improving developer experience, mentoring engineers, and helping teams make decisions that balance delivery speed with long-term maintainability.
Technical ownership across backend platforms, cloud architecture, frontend products, and engineering operations.
Tools I use to design, build, ship, and operate reliable software.
Engineering outcomes across asynchronous messaging, serverless performance, resilient recovery, and API optimization.
Designed an asynchronous notification service with channel-based selection and separate queues for high- and low-priority notifications, supporting priority-aware processing and workload isolation.
Designed an asynchronous service that selects notification channels and separates high- and low-priority workloads into dedicated queues. This architecture supports workload isolation and gives urgent notifications a distinct processing path.
Designed budget-management and reconciliation workflows using the Saga pattern to coordinate multi-step operations, manage failures, and support consistent recovery across distributed services.
Used Saga-based workflow orchestration for budget management and reconciliation, with explicit handling for failed steps and recovery actions. This approach helps keep distributed financial workflows traceable and recoverable.
Optimized AWS Lambda builds using esbuild bundling and tree-shaking, reducing deployment package size by 90% and cold-start latency by 70%.
Introduced optimized bundling and tree-shaking to trim unnecessary code from deployment artifacts. Resume-reported results: 90% smaller packages and 70% lower cold-start latency.
Architected a fault-tolerant dead-letter queue redrive platform using AWS Lambda Durable Execution, with filtering, transformation, retries, and failure handling.
Built a DLQ redrive workflow with filtering, transformation, retries, and failure recovery. The platform supports processing 50K+ failed messages per execution.
Improved frontend-facing API performance with caching and request orchestration, and built asynchronous processing systems handling 5M+ tasks per month.
Applied caching and request orchestration to reduce API latency by 35%. Separately, built queue-based asynchronous processing with idempotency, duplicate-request protection, retries, workload isolation, and failure recovery, handling 5M+ tasks monthly.
Implemented a WebRTC-based system supporting daily classes and student sessions across 6,500 centers, enabling real-time audio/video communication for learning operations.
Implemented WebRTC-based communication and signaling workflows for daily student classes across 6,500 centers. The system focused on enabling real-time sessions at scale for distributed learning locations.
Contributed to a vaccination-drive system handling millions of vaccination records, with HIPAA-conscious data processing and submission to the relevant government systems for governance and reporting.
Supported workflows to process millions of COVID-19 vaccination records with attention to HIPAA-related handling requirements and submission to the appropriate government systems for governance records.
Designed multi-region application deployment with health-aware routing and failover behavior, directing traffic away from an unavailable region to help maintain service availability.
Designed routing across regional deployments with health-aware decisions and a failover path so traffic can be directed to a healthy region during a regional outage. The routing strategy needs clear health signals, controlled failover behavior, and validation of regional origin availability.
Practical notes on backend engineering, distributed systems, architecture, and production trade-offs.
Explore my Medium profile for technical deep dives and lessons from building and operating software systems. The link opens my latest published writing.
Cloud certifications and a computer science foundation.
I’m interested in Staff and Lead engineering opportunities where architecture, product thinking, and hands-on execution come together.
+91 97829 70790 · Gurugram, India