Kaushal Bhatt

Regulated crypto wallet and custody platform — Java 21, Spring Boot, Kafka, AWS

• –Engineered internal developer platform services—including an annotation-based JWT/WebAuthn auth framework adopted across all 17 microservices and a zero-downtime centralized config server—eliminating redeploy dependencies platform-wide.
• –Architected and maintain the core AWS MSK (Kafka) event backbone connecting wallet, KYC/AML, and payment services with strict ordering and TLS encryption.
• –Shipped cross-chain transaction syncing and fiat payment flows; integrated Fireblocks custody and CoinMarketCap market data.
• –Raised reliability 25% with circuit breakers, failover design and Datadog/CloudWatch alerting; sustain 99.9% uptime on 24/7 on-call.

• –Cut the operations team's manual workload from 100% to 10% by shipping runtime document generation with summarisation on an LLM-integrated content platform.
• –Improved system availability 25% by leading the architectural decoupling of legacy SOAP services into a resilient, RESTful service-oriented system.
• –Held 95%+ automated test coverage with JUnit and Mockito, reducing production defects 20%.
• –Cut junior onboarding time 40% with a structured mentorship programme built for the account.

• –Architected a real-time event-streaming pipeline on Kafka and AWS MSK processing async transaction streams with sub-second latency.
• –Scaled elastically past 10,000 daily transactions by leading 6 engineers off a monolith onto AWS microservices with Kafka async processing.
• –Cut manual inspection 70% and raised fraud detection 55% with a real-time risk pipeline — XGBoost on AWS SageMaker over Kafka streams.
• –Reduced system latency 57% through query optimisation, asynchronous processing and Redis caching.
• –Halved deployment time with CI/CD pipelines; wrote 8 microservices end to end and ran 24/7 production on-call.

• –Handled 1M+ transactions daily with a Kafka-based distributed pricing engine centralising pricing across KSA warehouses.
• –Led migration of a Magento storefront to Java microservices; pricing accuracy and delivery routing lifted sales 15–38%.

• –Cut delivery time from 7 days to 3 by automating order synchronisation with third-party inventory over an event-driven Kafka architecture.
• –Removed 40% of the refund-operations workload with a refund-automation system driven by order state.
• –Recognised as Best Performer of the Year; received the "Knight Award" three times.