Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

Hi, I’m Sumana. I’m a software engineer who enjoys turning complex problems into reliable, scalable solutions. I thrive on building robust backends, modern microservices, and data-driven features that users rely on every day. Currently at American Express in New York, I’m focused on reducing latency, improving fault tolerance, and delivering insightful analytics dashboards. I love collaborating with cross-functional teams and exploring AI-assisted tooling to streamline analyst workflows and automate routine tasks.…Hi, I’m Sumana. I’m a software engineer who enjoys turning complex problems into reliable, scalable solutions. I thrive on building robust backends, modern microservices, and data-driven features that users rely on every day. Currently at American Express in New York, I’m focused on reducing latency, improving fault tolerance, and delivering insightful analytics dashboards. I love collaborating with cross-functional teams and exploring AI-assisted tooling to streamline analyst workflows and automate routine tasks.

Sumana

Web Developer, Full Stack Developer, Front-End Developer, +2





Hi, I’m Sumana. I’m a software engineer who enjoys turning complex problems into reliable, scalable solutions. I thrive on building robust backends, modern microservices, and data-driven features that users rely on every day. Currently at American Express in New York, I’m focused on reducing latency, improving fault tolerance, and delivering insightful analytics dashboards. I love collaborating with cross-functional teams and exploring AI-assisted tooling to streamline analyst workflows and automate routine tasks.…Hi, I’m Sumana. I’m a software engineer who enjoys turning complex problems into reliable, scalable solutions. I thrive on building robust backends, modern microservices, and data-driven features that users rely on every day. Currently at American Express in New York, I’m focused on reducing latency, improving fault tolerance, and delivering insightful analytics dashboards. I love collaborating with cross-functional teams and exploring AI-assisted tooling to streamline analyst workflows and automate routine tasks.

Available to hire

Hi, I’m Sumana. I’m a software engineer who enjoys turning complex problems into reliable, scalable solutions. I thrive on building robust backends, modern microservices, and data-driven features that users rely on every day.

Currently at American Express in New York, I’m focused on reducing latency, improving fault tolerance, and delivering insightful analytics dashboards. I love collaborating with cross-functional teams and exploring AI-assisted tooling to streamline analyst workflows and automate routine tasks.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Intermediate

Language

Javanese

Advanced

Work Experience

Full Stack Engineer at American Express

June 1, 2024 - Present

Led decomposition of a Java monolith into Spring Boot microservices, reducing cross-service failures and improving trace visibility with OpenTelemetry. Designed and built internal REST APIs on AWS to support high-throughput fraud workflows, achieving sub-100ms responses and production error rates below 0.1% during peak load. Reduced P95 latency by 25% for fraud-decision services processing 5M+ authorization requests/day by profiling JVM hot paths and optimizing Spring request handling. Built analyst-facing dashboards using React and Redux to surface Kafka-streamed real-time risk signals, cutting manual fraud investigations by 25% for 180+ daily users. Prototyped an LLM-assisted Ask Amex tool using Python and LangChain to evaluate response quality and latency, informing automation decisions for CRM summaries and analyst workflows.

Software Engineer at American Express

June 1, 2024 - Present

Reduced fraud-decision latency by 25% across 5M daily authorization checks by profiling JVM bottlenecks, tuning Spring Boot services, and implementing a circuit breaker using Resilience4j; validated via Datadog APM and Grafana dashboards. Supported migration of a legacy monolith to Spring Boot microservices to improve deployment consistency and reduce cross-service dependency failures as evidenced by OpenTelemetry traces. Developed React/Redux dashboards surfacing real-time fraud signals from Kafka streams, reducing manual investigations by 25% per analyst productivity logs in ServiceNow. Conducted Python/LangChain PoCs on Vertex AI to evaluate agent workflows for 'Ask Amex assistant', generating insights that reduced manual CRM handling. Optimized PostgreSQL/Redis-backed risk APIs, cutting P95 latency from 420ms to 290ms and lowering AWS EC2 costs by 10–12% as measured by Cost Explorer and CloudWatch. Strengthened release stability by expanding JUnit coverage and adding Cucumber BDD

Graduate Web Developer at Northwest Missouri University

January 1, 2024 - April 1, 2024

Built a multi-role academic analytics platform using HTML/CSS, JavaScript UI, Java APIs, Python pipeline, Firebase Auth/Firestore, enabling automated ingestion of 1,000+ student records with secure access for admins, faculty, and students. Cleaned and transformed CSV datasets with Python Pandas; created Seaborn and Matplotlib dashboards for grades, attendance, and progression trends, reducing manual spreadsheet work by 40%.

Full Stack Developer at Capgemini

February 1, 2020 - July 1, 2022

Developed a financial application UI module using React/Redux and Tailwind CSS, reducing UI defects by 15% and increasing session duration for User Profile Management. Resolved CORS and API integration issues between React UI and Spring Boot services, reducing client-side request failures and improving reliability. Optimized Spring Boot transaction services to handle 500+ concurrent requests with a 2-second SLA. Improved accuracy of financial data validation by optimizing PL/SQL test scripts, increasing defect detection across ETL pipelines by 30%. Investigated and resolved production issues using AppDynamics and ELK, helping maintain 99.9% uptime. Maintained uninterrupted data processing by providing on-call support for 10+ AWS-based data pipelines, achieving a 97% SLA resolution rate.