UJ Uma Nath Jha

Lead Backend / Platform Engineer

Building reliable Java and AWS platforms for high-throughput financial systems.

AWS Certified Solutions Architect Associate Java 21 Spring Boot Distributed Systems

Lead Backend Engineer and AWS Certified Solutions Architect with 10+ years across banking, fintech, cloud migration, and platform engineering. Currently at JPMorgan Chase, I lead infrastructure for a high-throughput offers platform at 3,000 peak TPS — with measurable wins: ~95% less writer load, 70%+ lower infrastructure cost, and a JPMC Innovation Award for a datasource routing pattern adopted across teams.

Email LinkedIn GitHub Resume PDF
Offer Platform Flow 150 ms SLA
Campaigns Offer metadata
APIs 10-12 services
Aurora 1.2B instances
Channels Branch + digital
3,000Peak TPS
~95%Writer load cut
70%+Cost reduction

Selected Impact

Production engineering with measurable outcomes.

Database Scalability

Transaction-aware datasource routing

Designed and productionized a routing pattern for Aurora PostgreSQL that sends read-only traffic to scalable Aurora replicas while keeping writes and mixed transactions on the writer instance.

Reduced writer load by ~95% and database-related API failures by ~99% during peak ingestion.

Cost Optimization

EKS to EMR Serverless migration

Led migration of ingestion jobs from EKS to EMR Serverless to better match workload execution patterns and reduce operational overhead.

Reduced ingestion infrastructure cost by more than 70%.

Cloud Architecture

Active-passive deployment

Architected active-passive and blue-green deployment architecture for microservices on AWS ECS Fargate using ALB/NLB and Route 53 failover routing.

Improved release safety and production resilience.

Reliability

PostgreSQL partition-swap fix

Fixed a production issue where partition attachment validation caused 4-5 minute blocking windows and API failures during ingestion.

Reduced attach/detach time to milliseconds through prevalidated partition-key constraints.

Cross-team Influence

Internal platform adoption

Drove adoption of the datasource routing pattern across multiple teams and secured acceptance into an internal Spring Boot platform library.

Recognized with a JPMC Innovation Award.

Case Study

Routing reads around the writer at 3,000 peak TPS.

How a transaction-aware datasource routing pattern took the load off a single Aurora PostgreSQL writer without risking consistency.

Problem

During peak offer ingestion, the Aurora PostgreSQL writer became the bottleneck. Read-heavy traffic and mixed read/write transactions competed on one writer instance, driving up writer load and triggering database-related API failures at the worst possible time.

Approach

Designed a transaction-aware dynamic datasource routing pattern: read-only traffic is routed to scalable Aurora replicas, while writes and mixed transactions stay pinned to the writer to preserve transaction consistency and ordering semantics.

Results

Writer load fell by ~95% and database-related API failures by ~99% during peak ingestion. The pattern was adopted across teams, accepted into an internal Spring Boot platform library, and recognized with a JPMC Innovation Award.

A public example of my read/write routing approach for multi-datasource Java applications: github.com/jhaumanath/two-data-source

Experience

Backend, infrastructure, and platform leadership.

JPMorgan Chase

Senior Software Engineer (SSE-3)

August 2024 - Present
  • Lead infrastructure and backend platform initiatives for the Chase Bank Offers platform — 10-12 core APIs, Aurora PostgreSQL Global Database across 2 regions, ~1.2B personalized offer/customer instances, and 3,000 peak TPS at a 150 ms SLA.
  • Architected active-passive and blue-green deployment architecture for microservices on AWS ECS Fargate, using ALB/NLB and Route 53 failover routing to improve release safety and production resilience.
  • Designed and productionized a transaction-aware dynamic datasource routing pattern for Aurora PostgreSQL, routing read-only traffic to scalable replicas while keeping writes and mixed transactions on the writer instance — reducing writer load by ~95% and database-related API failures by ~99% during peak ingestion.
  • Drove cross-team adoption of the datasource routing pattern and secured acceptance into an internal Spring Boot platform library; recognized with a JPMC Innovation Award.
  • Led migration of ingestion jobs from EKS to EMR Serverless, reducing infrastructure cost by more than 70% while improving operational efficiency.
  • Fixed a critical PostgreSQL partition-swap issue where partition attachment validation caused 4-5 minute blocking windows; reduced attach/detach time to milliseconds with prevalidated partition-key check constraints.
  • Leading the design and build of a team-level Agentic AI platform using RAG, MCP-based tool integration, enterprise knowledge retrieval, auditability, and compliance-aware access controls for cross-service debugging, offer testing, cost insights, and infrastructure-upgrade planning.
  • Mentor interns and engineers through technical design reviews, implementation guidance, debugging support, code reviews, and production-readiness reviews.

Newgen Digital Works

Technical Lead, AWS DevOps

February 2020 - July 2024
  • Designed and built a Health Monitoring platform for 100+ servers and 150+ applications using Icinga2, Spring Boot agents, RabbitMQ, and Python plugins.
  • Reduced noisy primary email alerts from 100+ per day to ~15 actionable alerts per day through event filtering, aggregation, and notification workflow improvements.
  • Migrated and containerized monitoring workloads and 5 related projects to AWS using ECS, RDS, OpenSearch, and AmazonMQ.

ThoughtClan Technologies

Lead Backend Developer

November 2016 - February 2020
  • Built Spring Boot services and workflow components for tenant-based inspection and observation workflows.
  • Led backend development for MS SQL Server to MongoDB migration workflows and Activiti workflow automation.
  • Implemented SAP Hybris Marketing Cloud integration services using Java, Spring Boot, REST, OData, MySQL, and AWS.

Skills

Focused on backend systems that survive real production pressure.

Backend

Java 21, Spring Boot, REST APIs, Microservices, Distributed Systems, API Design, API Versioning

Cloud

AWS ECS Fargate, EMR Serverless, Lambda, Aurora PostgreSQL, DynamoDB, S3, SNS/SQS, API Gateway, Route 53, IAM

Data & Messaging

PostgreSQL, SQL Server, MongoDB, Redis, Kafka, RabbitMQ

Reliability

Datadog, Splunk, Dynatrace, CloudWatch, Icinga2, SLA/SLO Engineering, Load Testing, Production Debugging

Delivery

Docker, Kubernetes, Terraform, Jenkins, Spinnaker, Git, JUnit, Contract Testing, Integration Testing

AI Platform

Agentic AI, RAG, MCP-based tool integration, enterprise knowledge retrieval, auditability, compliance-aware access controls

About

Ten years of moving toward harder production problems.

I started at Cognizant in 2016 building AEM components, then led backend and workflow development at ThoughtClan, ran AWS DevOps and monitoring platforms at Newgen, and now lead infrastructure for one of Chase's highest-throughput offers platforms.

I gravitate toward systems under real production pressure — databases at the edge of their limits, deployments that need to be boring, and infrastructure costs that need to shrink without breaking anything. That's where the datasource routing pattern, the partition-swap fix, and the EKS-to-EMR migration came from: each was a production problem first, and a talk track second.

Outside the day job, I stay close to platform engineering, observability, and the fast-moving AI tooling space — including building agentic tooling that makes infrastructure teams faster.

10+ yearsBackend & platform engineering
3,000 TPSPeak throughput supported
1.2B+Offer/customer instances managed
70%+Infrastructure cost reduction delivered

Education & Certifications

Credentials that back the work.

B.Tech in Information Technology

Future Institute of Engineering and Management, Kolkata

2015

Contact

Open to Lead Backend and Platform Engineering roles.

jhaumanath93@gmail.com +91-9066533885 LinkedIn GitHub