Quantum technologies promise to solve practical problems in chemistry and material science with a level of precision and speed that is beyond the reach of any classical supercomputer. They also allow us to envision impregnable information networks for secure communication and provide us with the most precise and gentle probes to study delicate materials and biological samples.

Photonics is a strong contender for building these quantum technologies: unlike other physical substrates for quantum technologies, light does not decohere at room temperature, light qubits can be easily encoded in time bins, allowing for massive scalability, and there is already very significant industrial capacity built around it.

Different states of light can be used to encode qubits. In fact, there are many families of states of light that can be used to encode qubits, called bosonic code states, of which cat and grid (or Gottesman-Kitaev-Preskill) states are two famous examples.

We investigate how these states can be generated in optical settings. Their generation requires nonlinear interactions between modes of light, which can be mediated by matter in the form of nonlinear materials or by measurements on parts of an entangled (quantum-correlated) state using photon counting detectors.

We use mathematical and computational tools, including inverse design, machine learning and tensor networks, to obtain optimal designs for structures, devices and protocols needed for the generation of these states and the operation of the detectors needed to measure them. These tools are not only useful for designing states for quantum computing, but also for generating states of light with metrological applications, such as squeezed light in which vacuum noise in a quadrature of the electromagnetic field is suppressed. At the same time, bosonic-code states can provide a viable route to build a network where quantum, instead of classical, information is exchanged: a quantum internet!

Finally, we also investigate the possibilities and limitations offered by near-term quantum computers known as boson samplers. These are not universal quantum computers, but offer the possibility of building quantum hardware with very minimal requirements that can exponentially outperform the most powerful supercomputers at certain restricted tasks.

We collaborate with academic colleagues at Polytechnique and beyond, across Canada and also abroad, as well as with government and industrial partners.

All our papers are available on the arXiv. You can also browse them on Google scholar.

You can find recordings of previous talks given by group members at underline and youtube.