Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

Full-stack software engineer with 4+ years of experience building scalable APIs, distributed systems, and AI-powered enterprise applications across financial services. Experienced in integrating production AI capabilities such as RAG, LLM inference pipelines, and semantic search into backend services and React-based internal tools. Proven track record of improving system performance and reliability using query optimization, caching, Kafka-based real-time processing, and AWS (EKS, Lambda, DynamoDB). Strong background in CI/CD automation and observability to support platforms used by tens of thousands of daily users.…Full-stack software engineer with 4+ years of experience building scalable APIs, distributed systems, and AI-powered enterprise applications across financial services. Experienced in integrating production AI capabilities such as RAG, LLM inference pipelines, and semantic search into backend services and React-based internal tools. Proven track record of improving system performance and reliability using query optimization, caching, Kafka-based real-time processing, and AWS (EKS, Lambda, DynamoDB). Strong background in CI/CD automation and observability to support platforms used by tens of thousands of daily users.

Varun Duggirala

Full Stack Developer, Software Engineer, AI Engineer, +3





Full-stack software engineer with 4+ years of experience building scalable APIs, distributed systems, and AI-powered enterprise applications across financial services. Experienced in integrating production AI capabilities such as RAG, LLM inference pipelines, and semantic search into backend services and React-based internal tools. Proven track record of improving system performance and reliability using query optimization, caching, Kafka-based real-time processing, and AWS (EKS, Lambda, DynamoDB). Strong background in CI/CD automation and observability to support platforms used by tens of thousands of daily users.…Full-stack software engineer with 4+ years of experience building scalable APIs, distributed systems, and AI-powered enterprise applications across financial services. Experienced in integrating production AI capabilities such as RAG, LLM inference pipelines, and semantic search into backend services and React-based internal tools. Proven track record of improving system performance and reliability using query optimization, caching, Kafka-based real-time processing, and AWS (EKS, Lambda, DynamoDB). Strong background in CI/CD automation and observability to support platforms used by tens of thousands of daily users.

Available to hire

Full-stack software engineer with 4+ years of experience building scalable APIs, distributed systems, and AI-powered enterprise applications across financial services. Experienced in integrating production AI capabilities such as RAG, LLM inference pipelines, and semantic search into backend services and React-based internal tools.

Proven track record of improving system performance and reliability using query optimization, caching, Kafka-based real-time processing, and AWS (EKS, Lambda, DynamoDB). Strong background in CI/CD automation and observability to support platforms used by tens of thousands of daily users.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Intermediate

Work Experience

Software Development Engineer at Morgan Stanley – Enterprise AI Platform

July 1, 2025 - Present

Contributed to the core backend of Morgan Stanley’s internal AI platform for financial analysts. Built Python (FastAPI) microservices with RAG pipelines and LLM inference workflows supporting 20,000+ daily analyst interactions. Designed and owned a REST API layer with caching, async orchestration, and rate limiting to reduce average latency by 20% on retrieval endpoints. Architected a Pinecone-backed vector retrieval system over proprietary financial research corpora, improving retrieval relevance and reducing analyst-reported hallucinations via chunking and embedding parameter tuning. Built a React + TypeScript research intelligence dashboard for internal analysts with performance optimizations (memoization, lazy loading). Enforced OAuth2 and JWT security, used DynamoDB for metadata and session state, and deployed containerized services on AWS EKS to maintain scalable, low-latency operations. Added Pytest unit/integration tests for FastAPI and RAG components to maintain coverage and

Software Development Engineer | Morgan Stanley – Enterprise AI Platform at Morgan Stanley

July 1, 2025 - Present

Contributed to the core backend of Financial Research Document Intelligence, Morgan Stanley’s internal AI platform for financial analysts. Built Python (FastAPI) microservices implementing RAG pipelines and LLM inference workflows to support 20,000+ daily interactions. Owned the REST API layer with caching, async orchestration, and rate limiting, reducing average API latency by 20% on core retrieval endpoints. Architected a Pinecone-backed vector retrieval system over proprietary corpora, improving retrieval relevance and reducing hallucinations via optimized chunking and embedding parameters. Built a React + TypeScript research intelligence dashboard with memoization and lazy loading. Enforced OAuth 2.0 and JWT security across microservices, stored metadata/session state in DynamoDB, and deployed on AWS EKS. Wrote Pytest unit/integration tests for FastAPI services and RAG pipeline components.

Software Engineer at Capital One

December 1, 2023 - June 1, 2025

Redesigned Java (Spring Boot) backend services for credit card fraud detection and risk decisioning, improving analyst-facing API response times by 20%. Built TypeScript-first React dashboards integrating live fraud scoring signals, model outputs, and transaction data using Redux. Implemented OAuth 2.0 and JWT security and supported PCI-DSS readiness, including encryption in transit/at rest. Enabled real-time fraud scoring for 20,000+ daily card transactions via production inference APIs serving Scikit-learn models. Reduced query execution time by 35–40% through MySQL join/subquery/index optimization across fraud analytics endpoints handling millions of daily transactions. Led migration of 3 fraud services to AWS serverless (Lambda, S3), reducing deployment time from 40 to 25 minutes and rebuilding CI/CD with automated testing gates. Maintained 99.5% uptime using CloudWatch structured logging and distributed tracing, reducing mean issue resolution time by 20%. Authored GraphQL schema

Software Engineer at Freshworks

January 1, 2021 - July 1, 2022

Developed backend features and REST APIs using Java and Node.js for Freshdesk and Freshsales, supporting customer workflows used by millions globally. Built React UI components for Freshdesk modules and internal dashboards with server-side rendering (SSR) to improve load performance and support WCAG 2.1 accessibility and SEO. Reduced API response times by 20–25% by addressing bottlenecks in high-traffic ticketing and CRM endpoints via MySQL query optimization and service refactoring. Implemented Redis caching and improved MySQL queries to boost retrieval speed by 25–30%. Increased code stability and reduced post-release defects by 30% by expanding unit/integration/frontend test coverage with JUnit and Jest. Reduced manual effort by 40% by automating recurring data validation using Python. Improved event delivery reliability by building asynchronous backend components for real-time workflow triggers, reducing processing delays by 20%. Collaborated across frontend, backend, and platf