Ayushi Agrawal

About Ayushi Agrawal

Java Backend Engineer @ Salesforce | Distributed Systems | Microservices | Cloud | LLM/AI | High-Scale Backend ArchitectureI am a Backend Engineer focused on building scalable, fault-tolerant, cloud-native backend systems using Java, Spring Boot, Microservices, REST/gRPC, Kafka, Redis, Docker/Kubernetes, and modern observability tooling.At Salesforce, I engineer backend platforms that process large-scale unstructured data, power secure access workflows, and integrate LLM-based reasoning systems into production services. What I work on:• High-throughput PII/Presidio pipelines (250K+ spans)• Distributed microservices with REST/gRPC interfaces• LLM-powered access-decision engines• Policy enforcement & authorization layers• HA services with fault tolerance, retries, circuit breakers• Feature-flag rollouts & zero-downtime deployments• Observability: Splunk • Grafana • Kibana • Tracing Core Strengths:Java • Spring Boot • Distributed Systems • MicroservicesREST/gRPC • Kafka • Redis • CI/CD • Kubernetes • CloudLLM Integration • Prompt Engineering • API DesignSystem Design (HLD/LLD) • Performance OptimizationScalability • Fault Tolerance • Event-Driven ArchitecturePreviously:EatClub → Ruby on Rails, Angular, Kubernetes, CI/CDJ.P. Morgan (Quant) → Python pipelines, anomaly detection, financial data automation (80% manual effort reduction)I’m passionate about backend architecture, system design, cloud-native engineering, and building reliable production systems that scale.

Experience

1. Software Development Engineer (Backend)

   Salesforce

   Nov 2023 — Present · Bengaluru, IN

   As a Java Backend Engineer (SDE-II) at Salesforce, I design and build high-scale, distributed backend systems leveraging Java 11+, Spring Boot, Microservices, REST/gRPC, Kafka, Containerized Deployments, CI/CD, Cloud-Native Architecture, and LLM/AI-driven components. I contribute to the architecture, development, and reliability of enterprise-grade platforms that process large volumes of unstructured data and power compliance, access control, and automation workflows across Salesforce.I engineer high-throughput data processing pipelines integrated with Microsoft Presidio for advanced PII classification (250K+ spans), including asynchronous flows, parallel execution, secure data transformation, span-level masking, metadata extraction, and policy-driven access control. These pipelines operate as part of a distributed microservice ecosystem, optimized for low latency, fault tolerance, and horizontal scalability.I designed and implemented an LLM-powered access-decision and troubleshooting engine that performs context extraction, semantic search, rule retrieval, and AI-based reasoning using prompt-engineered templates. This system automates complex decision workflows, reducing manual investigation time by 80%+, and integrates with internal event-driven services, caching layers, and microservice APIs.I build and maintain core platform components:REST/gRPC APIsInterceptors, filters, middleware, and authN/authZ layersTag-based policy enforcement enginesHigh-availability (HA) microservicesEvent-driven and async processing modulesThese components ensure compliance across distributed systems and reduced unauthorized access incidents by 40%.I improved service reliability by integrating Splunk (logging), Grafana (metrics), Kibana (search), distributed tracing, and feature flags for safe deployments. I implemented circuit breakers, retry patterns, and rate limiting to strengthen resiliency.

Education

• Malaviya National Institute of Technology Jaipur

  Bachelor of Technology - BTech