At Xizam, I designed and deployed AI agent workflows with LangChain and LangGraph for retrieval-augmented generation and tool-calling. I also connected LLM APIs, vector databases, and internal tools in automation pipelines.
On my AI Support Agent project, I built a production support agent with grounded retrieval, order and subscription lookup tools, and conversational memory. I stress-tested it against scripted adversarial attacks and fixed a session-isolation vulnerability.

