Mani Shankar

• Architected and built a high-throughput, event-driven ingestion pipeline (Java, AWS SNS/SQS) processing
100, 000+ daily financial records with sub-second latency and high uptime.
• Built a parallel-processing cash-reconciliation service using AWS Lambda and Java multithreading, sustaining
10, 000+ concurrent requests at peak with sub-second response times.
• Engineered a hybrid transaction-prediction service integrating a Kotlin/Quarkus microservice backend with a
Python xgboost model, deployed for real-time inference on Amazon EKS for high availability.
• Designed an ML-based fund-matching service (bigram analysis, sentence embeddings) that reduced manual
reconciliation by ~40% across a large book of fund assets.
• Delivered GenAI-powered backend platforms — a client document-processing onboarding service and a
retrieval-augmented copilot (LangChain, Azure AI Search) — cutting client onboarding cycle time from 5 days
to under 1 day.
• Tuned RAG retrieval and caching to reduce LLM token cost ~35% and retrieval latency ~40%.
• Partnered across product, compliance, and engineering teams to drive backend automation delivery at
enterprise scale; contributor on a 12-engineer platform team.

• Led full-stack development of a network-management platform on Spring Boot microservices, delivering APIs
across 10+ services for 50, 000+ daily active users with high availability.
• Designed a Kafka-based event-streaming and async-messaging layer across 10+ microservices, enabling
real-time notifications and decoupled service communication at scale.
• Implemented an API gateway and identity management with Kong and Keycloak — centralizing SSO
authentication, enforcing rate limiting, and securing previously exposed internal APIs.
• Drove CI/CD and performance improvements across a team of 8, cutting deployment time from 4 hours to 45
minutes.
• Mentored 5 junior and mid-level engineers on distributed-systems design and microservices best practices,
and led backend architecture decisions for long-term scalability.
◦ Keycloak CI/CD PostgreSQL Java Docker

• Developed Java/Spring REST APIs processing 1 M+ monthly financial transactions across compliance-critical
banking workflows.
• Optimized a trade-data compliance reporting batch pipeline (Java, Hibernate, Informatica, Oracle), cutting
execution time from 5 hours to 20 minutes (~93%) via query tuning, index optimization, partition pruning, and
bulk PL/SQL processing.
• Led the transition from a monolith to modular microservices across 6 services for compliance-critical
workflows.
• Automated deployments via CI/CD, moving from monthly to weekly releases.
• Contributed to distributed transaction management and regulatory reporting workflows used by 500+ internal
stakeholders.
Tech Stack:
◦
◦
◦ APIs
◦ Docker

• Developed and maintained RESTful APIs for a multinational bank using Spring Boot and Hibernate,
supporting high-volume transaction workflows for 10, 000+ users across 3 countries.
• Integrated Apache Kafka for async messaging and event-driven notifications, improving system decoupling
and processing reliability.
• Contributed to migration from a monolith to service-based architecture across 4 core modules, reducing
incident recovery time by ~50%.
• Built Angular frontend modules integrated with backend APIs for banking-portal users across 3 countries.
Tech Stack: Java
◦ Spring Boot
◦ Hibernate
◦ Apache Kafka
◦ Angular
◦ REST APIs
◦ MySQL
◦ CI/CD