// Hello World — I'm
Senior Software Engineer · AI/ML · Cloud · Full Stack
Results-driven engineer with 4+ years building production-grade systems across AI/ML, Cloud, and Full Stack domains. Hands-on with GenAI solutions, LLM integration, RAG pipelines, and enterprise-scale microservices. Eligible to work in the US (STEM OPT).
// 01 — About Me
I'm Praneeth Koppolu, a Senior Software Engineer with 4+ years building production-grade systems across AI/ML, Cloud, Full Stack, and Data Engineering. From fine-tuning FLUX LoRA models and RAG pipelines to architecting microservices at American Express scale — I bring both depth and breadth to every problem. Currently based in the US on STEM OPT.
Outside of shipping code, I research GenAI applications in storytelling (MythVision), mentor students in VR/AR development, and stay sharp on the latest in LLMOps and cloud-native architectures.
// 02 — Projects
End-to-end GenAI application generating Indian mythology stories illustrated in a custom art style. Combines GPT-4o-mini for narrative generation with a fine-tuned FLUX LoRA model hosted on Replicate for image synthesis. Full LLMOps pipeline with prompt engineering and model versioning.
Python library and CLI that studies an existing backend's framework, folder conventions, naming, authentication, and validation patterns before generating REST API layers that fit the codebase. Supports deterministic offline generation and optional architecture-aware LLM generation with safe dry-run previews.
Serve transformer models on GPU using NVIDIA Triton Inference Server. Optimized for high-throughput inference with dynamic batching, model versioning, and performance benchmarking at scale.
End-to-end GenAI application generating Indian mythology stories illustrated in a custom art style. Combines GPT-4o-mini for narrative generation with a fine-tuned FLUX LoRA model hosted on Replicate for image synthesis. Full LLMOps pipeline with prompt engineering and model versioning.
Real-time data pipeline built with Kafka, Spark, and PostgreSQL using Docker. Streams, processes, and stores ride events in real time for analytics and insights. Production-grade architecture with fault tolerance and horizontal scaling.
Full-stack e-commerce web application with product catalog, cart management, user authentication, and payment integration. Built with modern web technologies and responsive design principles.
Train, log, deploy, and monitor an NLP sentiment analysis model like a production engineer. Full MLOps lifecycle from experiment tracking to Kubernetes deployment with FastAPI serving layer.
Real-time data pipeline built with Kafka, Spark, and PostgreSQL using Docker. Streams, processes, and stores ride events in real time for analytics and insights. Production-grade architecture with fault tolerance and horizontal scaling.
// 03 — Experience
// Certifications
// 04 — Tools & Technologies
// 05 — Skillset