3
0Welcome to the Innovator Coffee, a podcast that bridges the gap between people and the world of AI and innovation. Follow us to explore the top AI products, ecosystem insights, and the emerging trends.
In this episode of Innovative Coffee, host Wickey Wang and Tom are joined by three cutting-edge founders and AI leaders:
Yu Zhou, Professor at Columbia and founder of Arklex.ai
Quanlai Li, founder of ChatSlide.ai and former ML engineer at Uber and Robinhood
Neethan, founder of Hallidai, and previously with Alexa and TikTok monetization
Together, they explore the rapidly evolving world of AI agents, agent frameworks, and how 2025 is shaping up to be the "Year of the Agent." Topics include:
Defining the next generation of work with agent-based automation
Real-world applications in sales, design, e-commerce, and enterprise workflows
The rise of MCP (Modular Context Protocol) and Agent-to-Agent (A2A) architectures for interoperability
AI agents as “AI interns” or “chiefs of staff” that augment rather than replace human workflows
Key technical challenges including multimodal capabilities, scalability, data privacy, and hallucination control
A candid look into startup GTM strategies, personalization, and the future of self-improving agents
A summarized version of each section of the podcast with timestamps, with a total of fewer than 8 sections:
00:00 – Introductions & Founders' Backgrounds
Host Wickey and Tom welcomes Yu Zhou@Arklex.ai, Quanlai Li, founder of ChatSlide.ai and Neethan ,former ML engineer at Uber and Robinhood
03:30 – Defining Next-Gen Work with AI Agents
The panel defines "next-gen work" and explores how AI agents reduce cognitive load, automate workflows, and augment human productivity across enterprises.
10:00 – Technical Differentiation & Scalability Challenges
Zhou explains the backend and scaling advantage of Arklex.ai AI. Quanlai and Neethan highlight challenges in agent design, metric tracking, and visual context.
18:00 – Collaboration vs. Efficiency: AI's Role in Teams
A debate on whether AI agents enhance or replace collaboration, and how “AI interns” enable individuals while still keeping humans in the loop.
25:00 – Emerging Capabilities: Multimodality & Experience Learning
Exciting new trends including multimodal AI, reinforcement learning from human-agent interactions, and the rise of MCP and A2A protocols.
35:00 – Protocols: MCP vs. A2A and Future Architecture
Detailed discussion on the role of Modular Context Protocol (MCP) vs. Agent-to-Agent (A2A) in scaling agent communication and coordination.
45:00 – Security, Privacy, and Hallucination Mitigation
Tactics and tools to reduce hallucination, enforce data privacy, and secure AI workflows—especially in enterprise agent environments.
52:00 – The Ideal AI Teammate & Final Reflections
Panelists describe their dream AI teammate and share final insights on GTM strategy, agent personalization, and what they’d build if starting over.
Hosts:
Tom Kong
*Stanford EE alumni,
*Founder@ Stanford AGI Adventist Community (10K+ members so far from top VC, Engineers, startups from Silicon Valley )
*AI Lecturer, a serial entrepreneur in media and data. Advisor @ techtimes.com
*AI deployment for 8 years, with NLP and recent LLMs (RAG, Agent, Diffusion)
Wickey Wang
*IT Security Compliance/Auditor & CSTU Faculty
*Growth fund VC advisor, Early stage fund fellow and Angel Investor with cybersecurity and AI focus
*GAI Security book co-author
Special thanks to Hannah Wang who did the wonderful job to assist to complete this podcast. www.linkedin.com
Questions, Suggestions, Feedback and Comments? You can find us in LinkedIn:
