Lucidic AI

AI Agent Training via Simulations

Y CombinatorWinter 2025ActiveB2BAIOpsArtificial IntelligenceDeveloper ToolsSaaS

What they do

Problem

Enterprises struggle to make general-purpose frontier AI agents reliably understand and operate within their specific business processes without modifying model weights.

Solution

Lucidic AI provides a simulation-based training layer that adapts AI agents to a company’s real operating environment without changing the underlying model weights.

Who it's for

Enterprise teams deploying AI agents for internal operations and workflows.

Founders

Abhinav Sinha

Founder

Hi, I’m Abhinav the founder and CEO for Lucidic AI! I’m from Stanford where I studied Computer Science with an AI specialization (both bachelor's and master's) and did research at Stanford's AI Lab. I’ve worked at Apple as a software engineer and at Citadel Securities and Susquehanna International Group as a quant. In my free time, I like to play basketball, pickleball, lift, and rewatch Christopher Nolan movies! Feel free to reach out, would love to chat -- abhinav@lucidic.ai

Andy Liang

Founder

Founder/CTO at Lucidic AI, working on an analytics, testing, and simulations framework for AI agent devs. Stanford BS/MS Computer Science (AI specialization). Ex Citadel, AppLovin. Previously built GrocerCheck, a COVID-19 web app helping shoppers social distance effectively, with over 200,000 users and endorsed by the Government of Canada. Reach me at andy@lucidic.ai

Jeremy Tian

Founder

Hi! I'm Jeremy Tian, one of the founders and the Chief Scientist at Lucidic AI, working to explain your AI models' decisions. I have a BS/MS in Computer Science with an AI specialization from Stanford University. Ex. Quantitative Trader at DRW and software engineering/machine learning engineering at Steel Dynamics. Contact me at jeremy@lucidic.ai!