Rajesh Kumar

AI Engineer & Generative AI Application Developer · Chicago, IL · Open to relocation within US

LLM Applications · RAG · Multi-Agent Systems · Frontend Engineering · (773) 916-1071

rajeshdevandla11@gmail.com LinkedIn GitHub Portfolio

Professional Summary

AI Engineer with 5+ years of software engineering experience building production-grade backend systems, GenAI applications on AWS, and modern frontend interfaces. Experienced driving API/integration design, modernization, and technical strategy including work with MuleSoft and event-driven architectures. Specialized in designing and shipping LLM-powered products using Amazon Bedrock, RAG pipelines, multi-agent workflows, and conversational AI. Proficient in Java 17, Spring Boot, Groovy/Grails, Angular, Next.js, Vue.js, Python, and FastAPI. Actively targeting AI Engineer, Generative AI Engineer, LLM Application Engineer, and Agentic AI Engineer roles.

Technical Skills

Languages & Frameworks

Java 11/17, Spring Boot, Spring Data JPA, Hibernate, Grails, Groovy, REST APIs, Microservices, Angular, React, Next.js, Vue.js, Svelte/SvelteKit, TypeScript, JavaScript, Node.js, HTML/CSS, TailwindCSS, Python, SQL, FastAPI

Cloud & DevOps

AWS (EC2, S3, Lambda, SQS, Bedrock, SageMaker, CDK), MuleSoft, Event-Driven Architecture, Infrastructure as Code, Docker, Kafka, Jenkins, GitHub Actions, CI/CD Pipelines

AI Engineering

Amazon Bedrock, OpenAI API, GitHub Copilot (agentic development), Generative AI, LLM Integration & Application Patterns, RAG Pipelines, FAISS, Prompt & Context Engineering, Orchestration Patterns, Multi-Agent / Agentic Architecture, Model Evaluation, Evaluation, LLMOps, Data Pipelines, Real-Time Data Pipelines & Analytics

Databases & Tools

PostgreSQL, MySQL, Schema Design, Stored Procedures, Query Optimization, GraphQL, REST APIs, JUnit, Mockito, Git, GitHub, Jira, Confluence, Agile/Scrum

Governance & Compliance

AI Governance Frameworks, Ethical AI / Responsible AI Practices, Regulatory Compliance, Secure Model Deployment

Engineering Practices

Object-Oriented Design, Design Patterns, Enterprise Application Architecture, Automated Testing (E2E, Integration, Regression), Observability, Production Support, Legacy System Modernization, Root-Cause Analysis

Projects

AgentFlow — Multi-Agent Orchestration System

Python · FastAPI · Amazon Bedrock · Streamlit · Multi-Agent · LLM

Built a multi-agent orchestration system (Planner, Researcher, Summarizer agents) on Amazon Bedrock without LangChain. Implemented custom agent base class with long-term memory stubs, failure recovery, and 55-test evaluation suite covering 10 principles. Dockerized multi-stage build for API and frontend.

GitHub Repo View Details

AskDocs AI — RAG-Powered PDF Chat Assistant

Python · Amazon Bedrock · FAISS · FastAPI · Streamlit · Docker

GenAI document assistant that lets users upload PDFs and chat with them using natural language. Built with Amazon Bedrock (Claude) for LLM generation and FAISS for vector search. Dockerized FastAPI backend with Streamlit chat UI, full RAG pipeline with PDF parsing and orchestration layers.

GitHub Repo Live Demo

AI Document Intelligence Platform

Java · Spring Boot · Python OCR · OpenAI API · Kafka · React · Docker · Kubernetes · MIT License

End-to-end document processing platform — uploads PDFs, extracts text via Python OCR, sends content to an LLM API for structured data extraction, and displays results on a React dashboard. Spring Boot microservices, Kafka async processing, Cypress E2E tests, Kubernetes/Terraform infra. 125 commits, MIT licensed.

GitHub Repo View Details

SupportGPT — AI Customer Support Assistant

Python · Amazon Bedrock · Sentiment Analysis · FastAPI · FAQ Chatbot

AI-powered customer support assistant that handles common support questions automatically, detects frustrated customers via sentiment analysis, and escalates to human agents when needed. Integrates Amazon Bedrock for LLM response generation with FAQ knowledge base and ticket summarization.

GitHub Repo View Details

SQLGenie — Natural Language to SQL Generator

Python · Amazon Bedrock · FastAPI · SQL Safety Validation

Converts natural language questions into safe SQL queries using Amazon Bedrock. Safety validation layer blocks destructive queries (DELETE, DROP, UPDATE) and returns only SELECT statements. Users type questions in plain English and get production-safe SQL with query explanation.

GitHub Repo View Details

Work Experience

Software Engineer II (AI Focus)

Mutual of Omaha · Remote, Chicago IL · Jan 2025 – Present

Built end-to-end features across Angular, Spring Boot, and PostgreSQL using modular service-layer architecture, reducing cross-service bugs by ~30%
Built 10+ REST APIs using Spring Boot, Spring Data JPA, and Hibernate supporting high-traffic insurance workflows with 99.9% uptime
Developed Angular reactive forms and UI components for a customer-facing portal serving 50K+ policyholders
Reduced synchronous API load by 45% by migrating 3 critical event workflows to AWS Lambda and SQS for async processing
Automated build, test, and deployment cycles using Jenkins and GitHub Actions CI/CD pipelines, cutting release time from 2 days to 4 hours
Implemented Apache Kafka for async event-driven communication between 4 backend services, reducing message loss by 60%
Integrated OpenAI API with custom prompt engineering to automate internal data classification, saving ~15 hours/week of manual effort
Shipped a GenAI internal assistant using Amazon Bedrock with multi-turn conversation chaining for faster policy document lookup
Maintained 90%+ code coverage via JUnit and Mockito; led PR reviews across a 6-person engineering team
Shaped AI strategy and governance frameworks enabling intelligent automation and reliable LLMOps.
Built and iterated on generative AI proofs of concept, retrieval pipelines, and LLM application workflows from prototype through production.
Improved AI application quality through systematic evaluation, prompt engineering, and controlled experiments; implemented LLMOps practices for reliable delivery.
Collaborated with business stakeholders and enterprise architects to define the roadmap for AI and generative AI capabilities aligned to measurable outcomes.
Established standards for AI governance, ethical AI practices, and secure model deployment.
Mentored engineers on LLM adoption, context engineering, and orchestration patterns; provided reusable templates to accelerate scalable delivery.

Software Engineer

Mutual of Omaha · Remote, Chicago IL · Jan 2023 – Dec 2023

Architected and shipped a Next.js customer portal with SSR and SSG, improving page load time by ~40% and boosting SEO rankings
Built reusable component libraries using Vue.js and Svelte/SvelteKit for internal tooling dashboards, reducing UI development time by ~25%
Designed and consumed GraphQL APIs, replacing REST endpoints for 4 high-traffic data views, reducing over-fetching and improving response times by ~30%
Implemented TailwindCSS-based design systems across multiple products, standardizing UI styling and cutting CSS bundle size by ~35%
Developed Node.js middleware services for session management, API proxying, and server-side data aggregation, improving frontend performance
Collaborated with UX designers to deliver fully responsive, accessible interfaces achieving WCAG 2.1 AA compliance
Integrated frontend applications with Spring Boot microservices via REST and GraphQL, maintaining consistent data contracts
Advanced automated integration capabilities for producer onboarding and management systems, including MuleSoft-based domain APIs.
Expanded automated testing strategies — E2E, integration, and regression frameworks — to advance engineering quality practices.
Championed CI/CD, observability, performance optimization, and operational resilience across the team.

Backend Java Developer

MUFG Bank · Hyderabad, India · Feb 2020 – Aug 2022

Built and maintained 15+ REST APIs using Java 11 and Spring Boot for internal banking workflows processing 10K+ daily transactions
Designed SQL schemas, stored procedures, and optimized PostgreSQL/MySQL queries, reducing average query execution time by 40%
Integrated Apache Kafka for async event processing across 3 core banking services, cutting inter-service latency by ~35%
Built internal operations dashboards in vanilla JS that reduced manual reporting time by ~5 hours/week for 200+ operations staff