Research
Emulation Research
Our emulation approach is simple: make the software think it is running on the proper hardware, in a real environment. Except we control time, the processor, and all the sensors on board. This lets us connect physics simulators, radio simulations, remote hardware, and even connect remote controllers to the software. It is not a new idea, but there is still not a good solution, and it is an active area of research all over the world.
Our Work
Our team has contributed to most of the popular emulation projects including Unicorn Engine, QEMU, Styx Emulator, Penguin Rehosting framework, AVATAR2, and more. We regularly use and contribute to research in conferences like USENIX, NDSS, etc. to make our tools better, and our AI agents more effective and token efficient. Using our work we have been able to reduce the cost by orders of magnitude while retaining strict cybersecurity requirements and maintaining feature parity. We are actively investigating the limitations of LLMs in developing emulators for development and cybersecurity purposes, and are looking to publish our work soon.