Looking for a job

Sai Subhash Velivela

@saisubhashvelivela

Forward Deployed AI Engineer building production LLM tools at FedEx, including runbook search and customs HS-code classification.

United States

Message

At FedEx, I built Java-based AI tools integrating Large Language Models for runbook search and customs HS-code classification. These tools achieved a 96.3% LLM tool-call success rate and reduced on-call MTTR by 71%.

I also engineered Kafka event-driven systems and optimized Spring Boot APIs and AWS Lambda cold starts. That work improved message delivery reliability and reduced P99 latency and monthly compute costs.

At Charles Schwab Corporation, I built a real-time Kafka fraud detection pipeline on AWS integrating machine learning models, reducing incident detection time from 32 to 5 minutes. I also migrated customer records to PostgreSQL with zero data loss.

Earlier, at AXIS Clinicals Limited, I integrated AWS Bedrock services and LLMs for adverse-event narrative summarization, cutting medical-monitor query times from 4 hours to 22 minutes. At IKF Finance Limited, I worked on Java payment systems and event-streaming pipelines.

Experience

Work history, roles, and key accomplishments

Current

Java Full Stack Developer

Current

FedEx

Mar 2025 - Present (1 year 7 months)

Engineered a Kafka 3.6 event-driven architecture with 12 producers and 18 consumers on AWS EKS, achieving 99.97% message delivery reliability processing 2.3 million daily events. Optimized Spring Boot APIs and AWS Lambda cold-start latency by 78%, integrating workflow automation and cloud-native patterns to reduce P99 latency from 180 ms to 38 ms and cut monthly compute costs by 61%.

Kafka Amazon EKS Spring Boot AWS Lambda Kubernetes Docker Terraform Elasticsearch OpenTelemetry Java React Spring AI GitHub Actions CI CD Large Language Models Generative AI advanced analytics Workflow Automation

Java Full Stack Developer

Charles Schwab Corporation

Jul 2024 - Feb 2025 (7 months)

Implemented Java-based OAuth2 security with AES-256-GCM encryption and method-level RBAC, ensuring secure API access and compliance with enterprise security standards. Built a Kafka real-time fraud detection pipeline on AWS Cloud integrating machine learning models to reduce mean time to detect incidents from 32 to 5 minutes, enhancing system reliability.

Java OAuth AES 256 GCM RBAC Kafka AWS Machine Learning PostgreSQL Hibernate Flyway Angular Jenkins AI advanced analytics

Java Full Stack Developer

AXIS Clinicals Limited

Oct 2022 - Nov 2023 (1 year 1 month)

Designed a GCP-native AI data pipeline processing 18 million events monthly, applying systems thinking and workflow automation to ensure zero data-integrity violations in compliance-driven environments. Reduced ETL runtime by 88% using Spring Batch and SQL Server optimizations, improving performance on a 400-million-row dataset for large-scale transactional platforms.

GCP AI Data Pipelines Workflow Automation SpringBatch SQL Server Junit Pact React GraphQL Kotlin RxJS AWS Large Language Models Generative AI

Java Developer

IKF Finance Limited

Mar 2021 - Sep 2022 (1 year 6 months)

Processed ₹2.4 billion monthly payment transactions using Java and Spring Batch, achieving 99.94% settlement success and zero regulatory violations through reliable workflow automation. Engineered a Kafka-driven event-streaming pipeline with PostgreSQL integration and AI-enhanced analytics, reducing event-processing errors from 2.3% to 0.006% and improving data pipeline performance.