AI Tuesdays: Fireside on General Purpose Agents
AI agents are evolving from assistants that answer questions to systems that plan, use tools, and carry multi-step work to completion. Plenty of agents dazzle in a demo; far fewer can be trusted to run unattended in production. The bottleneck isn't intelligence anymore, but rather reliability.
But what does it take to build a truly general purpose AI agent, one that can work reliably across different tasks, users, and real-world environments?
Our next edition of AI Tuesdays is a candid fireside conversation with Umesh Kumar, co-founder & CEO of Runable, and Shourya Vir Jain, co-founder & CEO of RamAIn (YC W26) — two companies attacking the same problem from different directions.
Runable is building a general purpose AI agent that carries out end-to-end work across domains: building apps and websites, designing decks, generating videos, running research, and more, all from a single interface.
RamAIn is building AI agents aimed at enterprise operations: a general purpose agent that automates any UI task and operates any software the way a person would, clicking through web portals, legacy systems, and desktop apps with no need for APIs or integrations.
What they'll get into:
What actually makes an agent "general purpose", and where that claim breaks in the real world
Why agents ace demos but stall in production, and what closes the gap: evals, guardrails, error recovery, and knowing when to hand back to a human
The build under the hood: planning, tool use, memory, orchestration, browser use, hand-offs to subagents
Where durable value accrues as raw model capability commoditizes: product, workflow, data, trust
An honest read on the next 12 months for anyone building or deploying agents
Format
A sharp, opinionated fireside chat. Doors open at 6:30 PM; conversation starts at 6:45 PM sharp and runs for 60 minutes, including audience Q&A. Open discussion and networking follow.
Who it's for
AI founders, AI PMs, AI engineers, AI operators, AI researchers and anyone building with or thinking hard about AI agents.
Seats are intentionally limited to keep the room high-signal.
Apply to attend!
Moderated by Amrut Rajkarne, AI & Tech Investor at Nexus Venture Partners.