I'm a Product Manager building AI-powered, agent-based experiences at Netcore Unbxd — products that guide users through complex decisions in ways that feel natural, not robotic. I came up through marketing and content, which means I obsess over the gap between what users say they want and what they actually do. That lens shapes everything: from how I frame a discovery problem to how I define what a successful AI interaction looks like.
Work
Below are some of my recent projects. Please request for the complete walkthrough of the locked projects.
01
Modular AI Chatbot Platform
Built a progression of chatbot implementations exploring how conversational AI evolves from basic prototypes to feature-rich systems with memory, tools, and database integration.
02
Podcast RAG System with Vector Search
Built a production-grade RAG system implementing automated podcast processing, vector embeddings, and semantic search to enable conversational Q&A over podcast content.
03
Voice Assistant with Persistent Memory
Built a phone-based AI voice assistant implementing conversation memory persistence to explore the technical challenges of stateful conversational AI systems.
04
AI Networking Assistant with Semantic Matching
Built an AI-powered professional networking system combining voice interaction, vector search, and automated call initiation to facilitate professional connections.
05
Autonomous Job Search AI Agent
Built an autonomous AI agent implementing multi-step agentic workflows for automated job search, content analysis, and personalized recommendations.
06
MCP Workspace Automation Agent
Built an autonomous AI agent using Model Context Protocol to connect Jira, Slack, and Notion, enabling natural language commands to execute multi-step workflows across productivity platforms.
Thoughts & Ideas
My take on building AI products, navigating the PM-marketing overlap, and designing agent-based experiences that users actually trust.
Designing Agent-Based Experiences That Actually Help Users Finish Things
What Marketing Taught Me About Users That Most PMs Miss
Why Building AI Products Is Fundamentally Different
Experience
From content and marketing into product — building AI-powered experiences at the intersection of user intent, business goals, and intelligent systems.
AI should make the user feel smarter, not replaced — that's the only design constraint that matters.
Stack
The tools I use to build AI-powered products, understand user behavior, and ship iteratively.
in short
Building AI products where the model serves the user, not the other way around
Yamine skipped presentations and built real AI products. Want to know how?
Yamine Durai was part of the January 2026 cohort at Curious PM, alongside 13 other talented participants.
click to see more details
