Hi, my name is

Thivin Anandh

PhD Student

SciML | MLOps | CFD | FEM | HPC

A passionate researcher working on Scientific Machine Learning and High Performance Computing for fluid flow problems. I'm excited to develop and deploy efficient and scalable algorithms for fluid dynamics problems using modern ML and HPC techniques.

Resume
Quick Links
Education Experience Teaching Publications Projects Certifications Achievements Events

About Me

Currently a PhD student at the AIREX Lab at the Department of Computational and Data Sciences at Indian Institute of Science. My research interests include:

Scientific Machine Learning
Physics Informed Neural Networks
High Performance Computing
Finite Element Methods
Computational Fluid Dynamics

I have three years of industry experience working as a software engineer at Accenture. I have also worked on the development of ParMooN, our in-house FEM package written in C++, and I'm the lead developer of FastVPINNs, an efficient implementation of hp-VPINNs written in python. I am a Google Certified Tensorflow Developer. I am also working with Shell India Markets Private Limited and ITC for my research.

Skill Sets

C++ C Python Matlab Tensorflow Docker Github Actions deal-ii Gmsh Pointwise CUDA MPI OpenMP

Recent Highlights

Post Doc Offer

Received a post doc offer at Brown University from Prof. George Karniadakis

Read More
Best Poster Presentation

Got best poster presentation award at IGHASC held at Heidelberg, Germany for our work on Singularly Perturbed problems.

Read More
Paper Submitted

We submitted our paper on SUPG Stabilisation to arXiV.

Read More

Education

2018 - 2024
Ph.D in Computational and Data Sciences
Indian Institute of Science, Bangalore
Advisor: Prof. Sashikumaar Ganesan
For more details, please refer here.
  • Finite Element Methods: Theory and Algorithms (Grade A+)
  • Parallel Finite Element Computing (Grade A+)
  • Parallel Programming (Grade A+)
  • Computational Heat Transfer and Fluid Flow (Grade B+)
  • Keyboard player for IISc's Music Band Rhythmica - Video Link
  • Member of CDS Wellness Committee from 2022-2023
2011 - 2015
Bachelor of Engineering (Mechanical)
Anna University, Chennai
GPA: 8.6 out of 10
  • Gold medalist for overall best outgoing student for batch 2011 to 2015
  • NCC: Cadet Under Officer(CUO) and C-Certificate holder.

Experience

Software Engineer - Accenture
Jun 2015 - Jul 2017
I was responsible for developing automated testing routines for a financial product at First Data. During my tenure, I received the following awards:
  • Exide Innovator Award: Recognized for developing an automation routine that reduced man-hours by over 90% and saved the client $3,000 USD annually.
  • ACE - Gold Standard: The highest award at Accenture for excellence in quality and delivery.
Teaching Assistant - Indian Institute of Science
Sep 2020 - Dec 2024
More Info

I have served as a teaching assistant for more than five courses over eight occasions at the Indian Institute of Science. The courses I assisted with include:

  • Finite Element Methods: Theory and Algorithms
  • Numerical Methods
  • Introduction to Artificial Intelligence and Machine Learning
  • Machine Learning Operations (MLOps)

Publications

Improving hp-Variational Physics-Informed Neural Networks for Steady-State Convection-Dominated Problems

Thivin Anandh, Divij Ghose, Himanshu Jain, Pratham Sunkad, Sashikumaar Ganesan, Volker John

arXiv preprint, 2024

PDF arXiv GitHub Project
FastVPINNs: Tensor-Driven Acceleration of VPINNs for Complex Geometries

Thivin Anandh, Divij Ghose, Himanshu Jain, Sashikumaar Ganesan

arXiv preprint, 2024

PDF arXiv GitHub Project
FastVPINNs: An efficient tensor-based Python library for solving partial differential equations using hp-Variational Physics Informed Neural Networks

Thivin Anandh, Divij Ghose, Sashikumaar Ganesan

Journal of Open Source Software, 2024

PDF Journal GitHub Project
An efficient hp-Variational Physics Informed Neural Network framework for Incompressible Navier-Stokes equations

Thivin Anandh, Divij Ghose, Ankit Tyagi, Abhineet Gupta, Suranjan Sarkar, Sashikumaar Ganesan

arXiv preprint, 2024

PDF arXiv GitHub Project
On the choice of hyper-parameters of artificial neural networks for stabilized finite element schemes

Subodh M Joshi, Thivin Anandh, Bhanu Teja, Sashikumaar Ganesan

International Journal of Advances in Engineering Sciences and Applied Mathematics, 2021

PDF Journal
Ensemble forecast of COVID-19 in Karnataka for vulnerability assessment and policy interventions

Sashikumaar Ganesan, Deepak Subramani, Thivin Anandh, Divij Ghose, Giridhara R Babu

medRxiv preprint, 2021

PDF medRxiv
Fast and Efficient hp-Variational PINNS framework for solving the Incompressible Navier-Stokes equations

Thivin Anandh, Divij Ghose, Sashikumaar Ganesan

International Conference on Computational Fluid Dynamics (ICCFD-12), 2024, Kobe, Japan

Conference Proceedings
GPU ACCELERATED FEM BASED LAGRANGIAN PARTICLE TRACKING FRAMEWORK FOR HUMAN AIR PATHWAY

Thivin Anandh, Sashikumaar Ganesan

ParCFD-2024, 2024, University Club Bonn, Germany

Conference Proceedings Project Page
A Deep Learning Simulation Framework for Building Digital Twins of Wind Farms: Concepts and Roadmap

Subodh M Joshi, Thivin Anandh, Sashikumaar Ganesan

12th International Conference on Simulation and Modeling Methodologies, Technologies and Applications (SIMULTECH-2022), 2024, Lisbon, Portugal

Conference Proceedings Project Page
Computational Ship Hydrodynamics: Modeling and Simulation

Sashikumaar Ganesan, Bhanu Teja, Thivin Anandh

Computational Science and its Applications, 1st edition, 2020

Taylor & Francis

Link

Projects

Academic Projects

Collection of research projects completed during my PhD on CFD, FEM and Scientific Machine Learning.

View Projects

GitHub Projects

Additional github projects, including teaching materials, coursework, tools, and open-source contributions that reflect my passion for coding.

View GitHub Projects

Certifications

Tensorflow Developer Certificate

Google

Issued: Jan 2024

Credential ID: 92608005

View Credential

The Finite Element Method for Problems in Physics

University of Michigan

Issued: Jul 2020

Credential ID: MCXRSFRASFXD

View Credential

Deep Learning

NVIDIA Deep Learning Institute

Issued: Sep 2023

Credential ID: 123456789

View Credential

Accelerating Data Engineering Pipelines

NVIDIA Deep Learning Institute

Issued: Feb 2022

Credential ID: 123456789

View Credential

TensorFlow Developer Professional Certificate

DeepLearning.AI

Issued: May 2021

Credential ID: 221640EDEDAC

View Credential

Achievements & Awards

Best Poster Presentation Award (IGHASC)
Oct 2024
Best poster Presentation award at Indo-German conference on Hardware aware Scientific Computing (IGHSAC) held at Heidelberg Germany.
Best Presentation Award (AI/ML Track)
Feb 2024
FastVPINNs: "An Efficient and fast hp-Variational PINNs for large scale simulations" at EECS-2024. Organized by IISc Bangalore.
Best Presentation Award (AI/ML Track)
Feb 2023
FastVPINNs: "Large Scale - AI Augmented simulations for wind farms" at EECS-2023. Organized by IISc Bangalore.
Gold Medalist: Best Outgoing Student
2015
Awarded Gold Medal for obtaining overall best outgoing student for the batch 2011 to 2015 at KEC.

Get in Touch

Email me at: thivinanandh [at] gmail [dot] com