Person

I am a Data scientist with focus on text analytics. I work at a reinsurance company because I love the breath of topics I am exposed to here.

Get in touch: info [AT] manuel-doemer.ch

What I love about my job is the diversity of thought and technical expertise around me: on the same day I might work with credit risk analysts, software architects, casualty underwriters and have lunch with a colleague from asset management.
You can see this also reflected in my path within the company: I started off in 2015 as a corporate actuary developing costing models for the economic loadings in reinsurance premiums. Over time and with new projects coming my way I became more and more focused on text analytics. In 2017 I eventually made the move into a data science position. The opportunity to persue a career in line with my passions is extremely valuable to me.

During my time in Costing Methods and Analytical Services I worked on the following topics:

Text is the ubiquitous form of storing information in the insurance industry. Therefore, my current focus on text analytics fits very well with the company's effort to create value out of the vast amount of unstructured data.
My main projects could be labelled as process automation problems, which involve

We are a bunch of data scientists who organise the Natural Language Processing Zürich Meetup. Come and join us at one of our next events!

In my department data scientists drive solutions end-to-end - we also bring successful prototypes in production. Obviously, we have help from infrastructure and architecture teams, but containerisation continuous integration and deployment to cloud platforms are part of our job.
During my PhD in computational chemistry at the École Polytechnique Fédérale de Lausanne (EPFL) I worked on the following topics:

Low energy structure of small biomolecules in the gas phase

A collaboration with the group of Thomas Rizzo (EPFL), who recorded conformer-specific vibrational spectra of small protonated biomolecules in the gas phase at low temperature. Employing a combination of computational methods we determined the molecular structures giving rise to the experimental spectra. Furthermore, we used the high-resolution experimental benchmark data to asses various computational methods to predict the correct relative energetics of low energy structures.

Deriving improved classical force field parameters by QM/MM force matching

With the goal to increase the accuracy of classical molecular mechanics force fields we implemented the recently developed force-matching protocol for an automated parametrisation of biomolecular force fields from mixed quantum mechanics/molecular mechanics (QM/MM) reference calculations in the CPMD software package. Such a force field has an accuracy that is comparable to the QM/MM reference, but at the greatly reduced computational cost of the MM approach. We have applied this protocol to derive in situ FF parameters for the retinal chromophore in rhodopsin embedded in a lipid bilayer.
See the QM/MM force matching page at lcbc.epfl.ch – Research – QM/MM force matching

Parameterisation of repulsive potentials for the self-consistent charge density functional tight-binding (SCC-DFTB) method

We employed iterative Boltzmann inversion to derive repulsive potentials for SCC-DFTB. We used reference data at the DFT/PBE level to derive highly accurate parameters for liquid water at ambient conditions, a particularly challenging case for conventional SCC- DFTB. The newly determined parameters significantly improved the structural and dynamical properties of liquid water at the SCC-DFTB level.

Dispersion-corrected atom-centered potentials (DCACPs) for the halogens

DCACPs are a recently developed method to cure the failure of DFT methods within the generalised gradient approximation to describe dispersion interactions. We complemented the existing library of DCACP parameters by the halogens and compared the performance of various dispersion corrected DFT methods in reproducing high-level benchmark calculations on weakly bound prototype complexes involving halogen atoms. See the DCACP page at lcbc.epfl.ch – Research - DCACPs

During my master’s thesis (2008) I worked on molecular dynamics simulations in electronic excited states, under the supervision of Kim Baldridge (Zürich University) and Daniele Passerone at the federal institute for material science (EMPA): A Force Field Approach to Photochemical cis-trans Isomerization.