Orimoloye Marvelous

At FlowDryftAI, I architect and deploy custom AI agents and multi-agent systems for enterprise clients. I also work with stakeholders to scope AI integrations and address operational bottlenecks.

I build data pipelines and RAG frameworks for client applications, and establish CI/CD and observability workflows for production deployments. I’ve also developed A/B testing infrastructure to evaluate model performance and agent workflows.

On my E-commerce AI Agents Suite portfolio project, I built three agents for customer support, retention, and smart inventory using FastAPI, LangGraph, and Pinecone. The inventory agent uses time-series demand forecasting, and I added Langfuse monitoring to the retention pipeline.

My other projects include a LangGraph system for codebase onboarding and documentation-drift detection, a multi-agent contract review pipeline, and an AI-powered client onboarding workflow. Earlier, during my Data Science internship at Kadesh Solution, I designed and integrated recommender systems for a mobile tutoring application.