Seismologist| AI | Building AI systems for Earth and Beyond
Seismologist| AI | Building AI systems for Earth and Beyond
Hi, I'm Akash — a Seismologist and Machine Learning Practitioner.
I'm a PhD student at the University of Washington, where I explore the seismic signatures of earthquakes, landslides, and other surface events. My research blends geophysics with deep learning to build intelligent systems that classify and characterize Earth's vibrations.
Outside academia, I'm passionate about applying ML to real-world problems — from image classification and LLM fine-tuning to reinforcement learning. I love building impactful, clean, and efficient models that work in the wild.
This portfolio is a collection of my projects at the intersection of Earth science and AI. Dive in to explore how I turn raw data into actionable insights, one model at a time.
Artificial Intelligence/Machine Learning Projects
Analyzing car reviews with LLMs (2025)
💬 Sentiment analysis, Q&A, translation, and summarization using Hugging Face Transformers.
🛠 Skills: LLMs, NLP, Hugging Face, BLEU score, prompt engineering
Current/Past Research Projects
Cluster Analysis of velocity models in Trans-Hudson Bay
Advisor: Dr. Amy Gilligan
University of Aberdeen
Discrimination of Icequakes and Tectonic events using unsupervised Machine Learning
Advisor: Dr. Peter Voss,
Geological Survey of Greenland and Denmark
Deep Long Period Seismicity beneath Yellowstone
Advisor: Dr. Keith Koper,
University of Utah
Paper in progress for below.
EGU GA 2022 Abstract submitted on 'Cluster Analysis of velocity models in Trans-Hudson Bay'
I will be presenting 3 posters at AGU FM 2021!
I received the AGU's virtual travel meeting grant!
I gave a poster presentation at EPOS Nordic Seminar.
My abstract on 'Discrimination of Icequakes and Tectonic earthquakes using unsupervised machine learning model' was accepted to be presented at 'European Seismology Commision' General Assembly 2021
I am enrolled in Remote Online Sessions for Emerging Seismologists (ROSES) 2021
I am taking 'seismology skill building workshop' organized by IRIS
My abstract on 'Discrimination of tectonic and non tectonic seismicity using unsupervised Machine Learning' is accepted for presentation at IUGG-IAGA Joint Assembly 2021
Errors Introduced in Estimation of Surface Wave Phase and Group Velocities Due to Wrong Assumptions: An Assessment Using a Simple Model For Love Wave - https://doi.org/10.5194/egusphere-egu21-779
Book Chapter – Mukhopadhyay S, Kharita A. 2021. Seismic Waves and Their Effect on Structures, Chapter 8 in the edited book “Advances in Wave Dynamics” Ed. S. Chakraverty and P. Karunakaran, Publisher World Scientific Publisher – Accepted.
Tentative - Discrimination of tectonic and non tectonic seismicity using Unsupervised Machine Learning Model
Skills and Strengths
Coding languages - Python, Matlab, Shell-script, FORTRAN
Software :
Seismic data processing - Obspy and Seisan, Hermann's computer programs in Seismology, Seismic Analysis Code
Mapping - GMT, PyGMT, Cartopy and arcGIS
Operating Systems - Windows and Linux
ML Packages - Tensorflow and Scikitlearn
IRIS Products - IRIS Earthquake Browser, jMAseis, IRIS gmap, IRISfetch, Seiscode, IRIS Data services
Numerical Modellling - Syngine, AxisSem, ASPECT
Visualization - Paraview, GIMP, Inkscape
Well familiar with Linux programming
Well familiar with HPCs
Well familiar with Jupyter notebooks
Well familiar with Github
Well familiar with how to acquire, process and interpret passive seismic data.
Well familiar with Machine Learning
Additional courses taken -
1) Computers , waves and simulations - Numerical Modelling course on Coursera - delivered by Computational Seismologist Dr. Heiner Igel, LMU Germany
2) Seismology skill building workshop organized by IRIS (2021)
3) Remote Online Sessions for Emerging Seismologists (2021)