Cygnux Labs
An independent research lab and think tank helping humanity make a safe transition into a post-AGI world, by making autonomous systems verifiable and accountable.
Mission
Our mission is to help humanity make a safe transition into a post-AGI world. Autonomous AI systems will soon do much of the world's work: trading, writing software, running logistics and administration. That could solve problems humanity has carried for centuries. It also brings new risks: failures that spread at machine speed, agents that can be steered by anything they read, and institutions that can no longer see what is being done in their name.
We believe this transition needs infrastructure that does not exist yet. Today, people trust an AI system because they trust the company behind it, or because its output looks right. That will not hold when millions of agents act on their own. As an independent research lab and think tank, we work from first principles to replace assumed trust with evidence anyone can check, records that cannot be quietly rewritten, and tests of how agents behave over years rather than minutes.
We publish our code, data and findings openly. For research collaborations, institutional partnerships or investor queries, contact us at info@cygnuxlabs.com.
At a glance
Bravish Ghosh
Founder and Principal Architect
Bravish works on making AI agents verifiable, easier to monitor and safer to trust. He leads this research at Cygnux Labs.
Previously, he was a Research Assistant at the Cambridge Judge Business School and worked at Microsoft on Azure core infrastructure. He studied Computer Science at ITER alongside Data Science at IIT Madras. His earlier work includes multi-agent crypto trading, AVX2-accelerated NTT for homomorphic encryption, and WGAN-A2C portfolio optimization.
He is always happy to talk about agent oversight, AI evals, verifiable infrastructure and ambitious engineering problems.
An independent lab
Remote by default
Cygnux Labs is small on purpose. It builds instruments, runs them on real agents, and publishes what it finds, including the results that did not go the way we expected.
We are looking for people who want to work on these questions with us.
We publish in the open