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

Hi, I’m Narendra Nadh Thota, a software engineer specializing in backend systems and agentic AI platforms. I currently build LLM infrastructure on AWS at Morgan Stanley, serving 500+ concurrent sessions. I’ve architected multi-agent RAG systems, asynchronous microservices, and CI/CD pipelines across fintech and B2B SaaS, delivering end-to-end solutions with zero critical incidents. I’ve developed ReAct agents and LLM evaluation frameworks in production, reducing research turnaround and enabling faster insights, while driving latency improvements and shorter release cycles through robust deployment practices.…Hi, I’m Narendra Nadh Thota, a software engineer specializing in backend systems and agentic AI platforms. I currently build LLM infrastructure on AWS at Morgan Stanley, serving 500+ concurrent sessions. I’ve architected multi-agent RAG systems, asynchronous microservices, and CI/CD pipelines across fintech and B2B SaaS, delivering end-to-end solutions with zero critical incidents. I’ve developed ReAct agents and LLM evaluation frameworks in production, reducing research turnaround and enabling faster insights, while driving latency improvements and shorter release cycles through robust deployment practices.

Narendra Nadh Thota

Full Stack Developer, Back-End Developer, Web Developer, +2





Hi, I’m Narendra Nadh Thota, a software engineer specializing in backend systems and agentic AI platforms. I currently build LLM infrastructure on AWS at Morgan Stanley, serving 500+ concurrent sessions. I’ve architected multi-agent RAG systems, asynchronous microservices, and CI/CD pipelines across fintech and B2B SaaS, delivering end-to-end solutions with zero critical incidents. I’ve developed ReAct agents and LLM evaluation frameworks in production, reducing research turnaround and enabling faster insights, while driving latency improvements and shorter release cycles through robust deployment practices.…Hi, I’m Narendra Nadh Thota, a software engineer specializing in backend systems and agentic AI platforms. I currently build LLM infrastructure on AWS at Morgan Stanley, serving 500+ concurrent sessions. I’ve architected multi-agent RAG systems, asynchronous microservices, and CI/CD pipelines across fintech and B2B SaaS, delivering end-to-end solutions with zero critical incidents. I’ve developed ReAct agents and LLM evaluation frameworks in production, reducing research turnaround and enabling faster insights, while driving latency improvements and shorter release cycles through robust deployment practices.

Available to hire

Hi, I’m Narendra Nadh Thota, a software engineer specializing in backend systems and agentic AI platforms. I currently build LLM infrastructure on AWS at Morgan Stanley, serving 500+ concurrent sessions.

I’ve architected multi-agent RAG systems, asynchronous microservices, and CI/CD pipelines across fintech and B2B SaaS, delivering end-to-end solutions with zero critical incidents. I’ve developed ReAct agents and LLM evaluation frameworks in production, reducing research turnaround and enabling faster insights, while driving latency improvements and shorter release cycles through robust deployment practices.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Work Experience

Software Engineer – AI Platforms at Morgan Stanley

August 1, 2025 - Present

Architected end-to-end agentic AI platform on AWS (ECS + Lambda) using LangChain multi-agent orchestration and RAG pipelines over proprietary financial data, enabling streamlined research workflows. Built production-grade RAG system with LangChain + FAISS + S3, integrating OpenAI/Claude LLMs with domain-tuned prompting, achieving 88% retrieval accuracy on internal knowledge bases. Delivered three containerized microservices (Python/FastAPI) for ingestion, inference, and orchestration on Kubernetes (EKS), achieving 99.7% uptime. Designed real-time streaming inference pipeline with asynchronous Python, WebSockets, and Redis pub/sub, delivering sub-second LLM responses to 500+ concurrent sessions. Implemented CI/CD with GitHub Actions → ECR → EKS and blue-green deployments, cutting release cycles from 2 weeks to 1 day. Built an LLM evaluation framework using LangSmith and custom Python harness for hallucination detection, latency SLOs, and cost-per-query tracking across four providers

Associate Software Engineer at Informative Web Solutions

May 1, 2021 - December 1, 2023

Owned production-ready RESTful APIs in Python (FastAPI/Django) powering core product workflows, sustaining 10K+ daily requests with sub-200ms latency, while developing frontend interfaces using React and TypeScript. Led multi-tenant PostgreSQL backend with row-level security and schema isolation, onboarding 50+ enterprise clients without downtime. Reduced p99 latency from 800ms to 180ms by introducing Redis caching and optimizing queries, cutting DB load by 40%. Built asynchronous task processing with Celery + Redis handling 1M events monthly with zero message loss. Achieved 85% test coverage via pytest, lowering post-release production bugs by 30% over two quarters. Containerized full stack with Docker, reducing onboarding time from 2 days to 2 hours across 3 product teams. End-to-end ownership of Stripe, Twilio, and SendGrid integrations for billing and notifications to all paying customers.