The emergence of the “Lobster Agent” has given many people their first real sense of the potential of AI agents. However, behind the hype lie significant challenges in security, efficiency, and infrastructure setup.
What can enterprises actually do with AI agents? Where should they start? How will operational workflows change?
Join this AI Innovation & Technology Application Seminar to explore these questions — so you don’t have to learn everything through trial and error.
This seminar, titled “After the Lobster — The Next Step of AI Agents Entering the Enterprise”, brings together first-hand insights from academia, product development teams, and enterprise adopters.
We focus not on future visions, but on real-world practice — helping participants identify their next step in the rapidly evolving AI agent landscape.
The keynote sessions will be delivered by CTOs and developers deeply involved in AI agent research and evaluation. Drawing on representative systems such as Hermes Agent and Manus, the talks will analyze:
Design principles of current agent architectures
Capability boundaries of AI agents
Methods for agent evaluation (Agent Evaluation)
With both academic backgrounds and product implementation experience, the speakers will provide a deeper understanding of how modern AI agents actually work.
We are witnessing several emerging trends — the rapid rise of “Lobster Agent,” Manus, Hermes, and other agent products has pushed Agentic AI from concept to deployable reality. At the same time, a growing gap has emerged between enterprise expectations and real-world implementation outcomes, where evaluation has become a critical yet often underestimated factor in success.
This panel will explore:
Which enterprise workflows are truly suitable entry points for AI agents?
How can organizations build evaluation mechanisms to measure agent effectiveness?
Why are organizational and process barriers often harder to solve than technical ones?
What does the viral success of the Lobster Agent signal for enterprise AI transformation?
Participants will leave with:
A framework for prioritizing AI agent adoption in enterprise environments
Practical design approaches for evaluation systems
Real-world lessons learned from both product teams and enterprise implementations
A clearer understanding of Agentic AI trends and direction
※ Topics are subject to change depending on speaker availability.
Enterprise technology leaders evaluating AI agent adoption
— Those who already recognize AI potential and want to prioritize use cases based on real-world experiments, avoiding unnecessary trial and error.
Developers and AI product owners in startups
— Teams building or refining agent-based applications, looking to learn from design approaches and pitfalls of products like Hermes and Manus.
Digital transformation and innovation teams in traditional industries
— Non-technical stakeholders who need concrete examples and evidence to communicate AI agent value and risks internally.
Managers responsible for AI procurement or vendor selection
— Decision-makers seeking a structured evaluation framework to compare the increasingly diverse AI agent solutions on the market.
Researchers and media covering the AI agent ecosystem
— Those tracking how AI agents transition from lab experiments to enterprise deployment, looking for grounded insights and industry perspectives.
This forum welcomes participants from across industries interested in enterprise AI agent applications and provides opportunities for cross-domain exchange and collaboration.