Led development of AI-driven fashion platform leveraging generative AI, LLM pipelines, and computer vision, achieving 40% user engagement increase.

Architected scalable Python-based ML pipelines using diffusion models for high-fidelity fashion visual generation, deployed on AWS infrastructure.

Designed and maintained RESTful microservices using FastAPI integrated with AWS Lambda and S3 for seamless data flow and integration across systems.

Build production-grade LLM applications, RAG systems, conversational AI, and intelligent recommendation platforms using GPT models, LangChain, LangGraph, vector databases, and modern ML architectures.

Design multi-agent workflows and AI automation systems that handle complex business processes, decision-making, tool orchestration, and end-to-end task execution.

Deploy scalable AI/ML solutions with Python and FastAPI across AWS, Kubernetes/EKS, CI/CD, APIs, and cloud-native infrastructure, with a strong focus on reliability and production performance.

Engineered PyTorch-based voice cloning system with 15 voice agents; fine-tuned Hugging Face models on 80 min/voice achieving 92% similarity, processing 1,000+ daily requests with <2s latency using multi-threaded audio generation pipelines.

Built real-time CCTV surveillance system with YOLOv8 and multi-camera tracking using Python multi-processing pipelines, reducing incident response time by 60%.

Developed YOLO-based malaria detection and blood cell classification system achieving 95% accuracy, processing 500+ daily images with automated Pandas-based data transformation pipelines.