.gif)
GenAI Engineer at Citi, building AI-driven solutions within institutional risk technology. Computer Science graduate from Northeastern University (May 2026), focused on ML engineering, distributed systems, and production-grade backend development. I'm drawn to problems where reliability matters and systems need to scale.
On the AI side, I've built CaduceusAI, a three-tier medical AI platform with local LLM inference, AES-256 PHI encryption, audit logging, and a LoRA fine-tuning feedback loop, and OmniRAG, a production RAG pipeline using LlamaIndex, LanceDB, BGE cross-encoder reranking, and Ollama for fully self-hosted document intelligence. On the distributed systems side, I've implemented Raft consensus, deployed multi-service architectures on AWS ECS Fargate with Terraform, and built circuit breaker, leader-follower, and leaderless quorum patterns from scratch.
The code I write tends to be modular, observable, and built for failure. I think about caching strategies and invalidation, graceful degradation when dependencies go down, schema migrations that don't break running services, and security boundaries that don't get bolted on at the end. Whether it's a FastAPI service, a Go backend, or a Next.js frontend, I try to write things that a teammate could pick up and extend without needing a walkthrough.
I work primarily in Python, Go, and Java, with experience across FastAPI, Docker, PostgreSQL, Redis, AWS, and Terraform.
View my website: https://pranavvis.tech
