Hey 👋, I'm Manu

     

My background in a nutshell.

---

• I am a computational scientist with expertise in numerical analysis, high performance computing, and scientific machine learning. I am interested in application of computational and stastistical learning methods to a wide range of scientific fields including natural sciences, geosciences, physical systems etc. I currently work as a Quantitative Research Analyst for the Citigroup's Credit Quant group in NYC.
• I have experience in developing and implementing novel algorithms to solve multiphysics CFD problems, using both data-driven deep learning techniques and classical mixed finite element methods (FEM).
• Currently, I work as a credit quant specializing in pricing and risk management of credit derrivatives in the risk-neutral framework, specifically in LATAM emerging markets.
• I love working on new challenging models and implementing them, while collaborating with others.
• Always looking for new collaborators and interesting projects.

Languages I work with:

• Python for general purpose programming and ML software development.
• C++ for high performance scientific computing and simulations.

Mathematical Skills:

• Numerical Analysis.
• HPC Parallel computing.
• Advanced Probability, Statistics.
• ML,DL.
• Biocomputing.
• Math Finance and Stochastic Calculus.
• Risk Neutral Pricing and Hedging of Credit derrivatives.
• Bonds, CDS, XCCY swaps.
• CFD Computational Fluid Dynamics.

Software packages I currently use:

• ML/ Data Science: Numpy, Pandas, Jupyter, Keras-Tensorflow2
• Visualization: Matplotlib, ParaView, gnuplot.
• Scientific computing: deal.II, FreeFem++, phoenix.
• My favorite editors and IDEs: Eclipse, Jupyter notebook, Colab, Emacs.

Languages and packages I used to work with:

• Fortran, Matlab, PyTorch, Scikit-learn

Checkout some of my recent projects and preprints:

---

Machine Learning:

1. Fluidlearn: A python based package to solve fluid flow PDEs using deep learning techniques.
2. Hands on practical ML projects:
   • Keras-TensorFlow2 based examples
   • PyTorch based examples
   • Scikit-learn based examples

High performance scientific computing:

1. [Space-time-DD](https://github.com/mjayadharan/MMMFE-ST-DD: A C++ based fluid flow simulator using multiscale space-time domain.
2. Poroelastic flow simulator: C++ based poroelastic fluid flow simulator using MPI.

Other open-source contributions:

1. FEM package deal.II: Most of the HPC packages I have written uses deal.II and I am also one of the contributors to this popular open-source FEM package.

Recent preprint:

1. Parallel computations to solve poroelastic flow: M. Jayadharan, E. Khattatov, I. Yotov, Domain decomposition and partitioning methods for mixed finite element discretization of the Biot system of poroelasticity, arxiv math.NA, 2010.15353.

Thank You-🙏🏼