Kam il to porek

Directed a customer-facing D2C service assistant engagement and shaped agentic verification experiences, integrating FastAPI services to move customers toward scheduling and payments. Built and implemented agentic verification AI agents for a D2C business using Luna conversion, translating client requirements into UX/UI architecture and reusable front-end components, and designing stable customer-journey flows (calendar, payments, CRM, messaging). Built event-driven integrations (GraphQL, gRPC, FastAPI) and implemented retrieval-augmented generation flows with improved answer support by about 28%. Established secrets management and automation for engineers, coordinating delivery via CI/CD, Docker, event-driven systems, Git, Jenkins, and WebSockets.

Developed an AI healthcare assistant (“iX Hello”) backed by FastAPI and LangChain, applying machine learning and LangChain to recommend care pathways and capture structured follow-up details. Converted clinical requirements into UX/UI state flows and mapped them to GraphQL schemas. Designed event-driven integrations using GraphQL, gRPC, and Redis, with evaluation and retrieval logic that reduced unsupported answers by about 28% prior to production. Implemented secrets management, automated testing workflows, and improved engineering coordination for reliable delivery.

Owned IoT platform technical direction including TypeScript services, WebSockets, and FastAPI APIs delivering device data for vehicles and industrial equipment. Established API service standards and reusable provisioning components to reduce device onboarding time. Selected WebSockets and event-driven pub/sub patterns for near real-time data exchange. Delivered data visualization for municipal and industrial use cases, improved AWS reliability with Redis caching, deployment checks, and structured observability to reduce recurring issues, and mentored engineers on API contracts and test coordination.