Senior Software Engineer AI, Python & Java Expert

Kevin Allen Senior Software Engineer Python, Java & AI Specialist

850-***-**** *****.*****.********@*****.*** Miami Beach, FL LinkedIn SUMMARY

Senior Software Engineer with 10+ years of expertise in architecting high-scale distributed systems and AI- driven automation. Specialized in Python (FastAPI, LangChain) and Java (Spring Boot, Microservices) for FinTech applications. Proven track record of leading the integration of LLMs and Generative AI into enterprise workflows at Amazon and Workday.

TECHNICAL SKILLS

• Languages: Python (Expert), Java (Expert), TypeScript, SQL (PostgreSQL/MySQL), NoSQL

(DynamoDB/Redis).

• Generative AI & ML: LLM Orchestration (LangChain/LangGraph), Amazon Bedrock, OpenAI API, RAG Architecture, Vector Databases (Pinecone, Milvus), PyTorch, Scikit-learn, Amazon SageMaker, Computer Vision (OCR).

• Backend & Architecture: Spring Boot, FastAPI, Microservices, Event-Driven Architecture (EDA), Distributed Systems, High-Throughput & Low-Latency Design, RESTful & gRPC APIs, Hexagonal Architecture.

• Data Engineering: Apache Spark, AWS Kinesis, Apache Kafka, ETL Pipeline Design, Data Modeling, Workday Prism Analytics.

• Cloud & Infrastructure: AWS (Lambda, ECS, S3, RDS, SNS/SQS, CloudWatch), Terraform, AWS CDK, Docker, Kubernetes, CI/CD (GitHub Actions, Jenkins, ArgoCD).

• Tools & Methodologies: Test-Driven Development (TDD), Domain-Driven Design (DDD), Agile/Scrum, Git, OAuth2/OIDC, SAML 2.0.

PROFESSIONAL EXPERIENCE

AMAZON Software Engineer II Dallas, TX Feb 2024 – Jan 2026

• Engineered an async multi-agentic orchestration layer using Python (FastAPI) and Amazon Bedrock, utilizing LangGraph to automate 85% of unstructured global invoice processing.

• Optimized LLM inference costs by 40% via a Java-based hybrid router that utilized deterministic OCR for standard templates while routing high-entropy exceptions to Bedrock.

• Architected a distributed anomaly detection system in Java (Spring Boot) via AWS Kinesis, processing 1.5M+ daily transactions with under 200ms p99 latency to block duplicate payments.

• Developed a Python-based RAG pipeline using Pinecone to enforce real-time tax compliance during data extraction, achieving 99.8% accuracy across 12 international jurisdictions.

• Automated multi-region infrastructure deployment using AWS CDK (Java) and Terraform, implementing granular observability for model drift and latency via custom Python/CloudWatch metrics.

• Led the transition to Hexagonal Architecture for FinTech microservices, reducing production hotfixes by 35% through standardized Java/Python TDD patterns and rigorous code reviews. WORKDAY Software Engineer Atlanta, GA Aug 2023 – Feb 2024

• Architected an Autonomous Reconciliation Engine using Java and Workday Orchestrate on the Workday Extend platform to automate cross-border payroll audits, reducing month-end close cycles from 8 days to 4 days.

• Integrated Python-based ML models (Scikit-learn) via high-throughput REST APIs to predict and flag high-risk journal entry discrepancies, resulting in an 80% reduction in payroll system failures for Fortune 500 clients.

• Engineered real-time data validation pipelines using Workday Prism Analytics and Python, improving financial reporting accuracy by 30% across 50+ enterprise production environments. AMAZON Software Engineer II Dallas, TX Aug 2022 – Mar 2023

• Refactored the core Java/Spring Boot ingestion engine for Global Accounts Payable, migrating to an event-driven architecture using AWS SQS/SNS and increasing throughput by 3x.

• Engineered a standardized Python validation framework, utilizing Pydantic to enforce schema integrity across 50+ internal microservices, reducing integration-related production bugs by 40%.

• Optimized AWS Lambda and DynamoDB query patterns, maintaining 99.99% availability and sub- 50ms latency during Prime Day peak traffic volumes. GUIDESTONE FINANCIAL RESOURCES Software Engineer Dallas, TX May 2016 –Sept 2021

• Refactored a legacy monolithic pension management system into a high-availability Java/Spring Boot microservices architecture on AWS ECS, utilizing Docker for containerization and improving system reliability to 99.99%.

• Engineered distributed ETL pipelines using Python and Apache Spark to perform complex financial reconciliations on $500M+ in assets, implementing multi-stage data validation to eliminate transaction discrepancies.

• Implemented a high-performance API Gateway in Java, integrating SAML 2.0 and OAuth2/OpenID Connect for secure SSO authentication across 200,000+ participant accounts while reducing auth latency by 35%.

• Optimized relational database performance in PostgreSQL by refactoring complex query patterns and implementing Redis-based caching for frequently accessed portfolio data, increasing transaction throughput by 40%.

• Developed a thread-safe financial calculation engine in Java, utilizing BigDecimal arithmetic for precise compound interest and tax withholding logic, ensuring 100% compliance with federal financial regulations.

EDUCATION

Bachelor’s degree Computer Science Rutgers University 2012 – 2016

Contact this candidate