---
title: "Anshuman Choudhury, PhD - Scientific Computing & AI Research in Bangalore Urban | Tal"
description: "Meet Anshuman Choudhury, PhD, Computational Materials Scientist (PhD) | Dislocation Dynamics & MD Simulation | AI/ML for Scientific Computing | Open to Opportunities based in Bangalore Urban. AI stack includes claude, codex, and github copilot. Explore public work links including github."
image: "https://tal.club/meet/anshuman-choudhury/share-image"
---

# Anshuman Choudhury, PhD on Tal

## Headline

Computational Materials Scientist \(PhD\) \| Dislocation Dynamics & MD Simulation \| AI/ML for Scientific Computing \| Open to Opportunities

## About

Computational materials scientist with a PhD from Université Paris\-Saclay and co\-authored research published in Nature Materials\. My work has focused on large\-scale atomistic simulations, scientific computing, and engineering systems across materials and nuclear science\. Currently building AI\-driven scientific computing workflows, surrogate modeling, and simulation pipelines aimed at accelerating engineering research and reducing computational cost\.

## Location

Bangalore Urban

## AI stack

claude

codex

github copilot

notion

claudecode

ChatGPT

## Work experience

Scientific Computing & AI Research at Self\-Employed, Oct 2022 \- Present · 4 yrs 1 mo, Bengaluru, Karnataka, India, • Physics vs black\-box ML for alloy strength: built an open benchmark on 444 multi\-principal\-element\-alloy records \(Borg et al\. 2020, public data\)\. Five dislocation\-theory features \(size misfit, modulus mismatch, VEC, shear modulus, Varvenne–Curtin scaling\) matched a 26\-element composition model on unseen alloy families \(R² 0\.50 for both\), and random train/test splits overstated accuracy by 23–31%\. Python, scikit\-learn, Random Forest, GroupKFold cross\-validation\. • Supported the Nature Materials \(2024\) paper through to publication: figures and manuscript discussions\. • Building Python surrogate models for materials\-property prediction on open datasets \(Random Forest, matminer features, Matbench benchmarks\), with code on GitHub\. • Machine Learning Specialization \(Stanford / Coursera, Andrew Ng\): Course 1 completed with certificate\.

Operations Lead at Family Business, May 2020 \- Jun 2022 · 2 yrs 2 mos, Took on operational leadership of a family trading business \(tea, Assam silk\) during a period requiring my return to India\. Modernized legacy cash\-based operations into digital payment systems and formal documentation processes; managed vendor relationships and business continuity through a volatile period\.

Doctoral Researcher at CEA, Jan 2016 \- Sep 2019 · 3 yrs 9 mos, Greater Paris Metropolitan Region, Built and ran atomistic simulations \(custom serial Fortran MD/NEB, EAM potentials\) of screw\-dislocation glide in bcc Fe and W; second author, Nature Materials 23, 47–51 \(2024\)\. Designed a multiscale NEB→MD pipeline: a 50\-image saddle\-point kink embedded in a \~1\.5M\-atom extended cell, so MD starts from the activated state instead of waiting for a rare nucleation event; reached dislocation lengths comparable to in\-situ TEM\. Identified a heat\-assisted mechanism: travelling kinks release kinetic energy that nucleates secondary kink pairs, giving long jumps below the Peierls stress; reproduced across three EAM potentials in Fe\. Ran a parametric study over potential, metal \(Fe vs W\), stress and temperature, with ensembles of 40 MD runs per condition: jerky glide is easier and more temperature\-sensitive in Fe; compressive stress plus temperature lowered the onset stress from 600 to 450 MPa \(one potential\)\.

Project Intern at Institute For Plasma Research, Jan 2015 \- Jun 2015 · 6 mos, Gandhinagar, India, Worked on neutron transport and flux analysis for Lead assemblies used in fusion reactor breeding blanket studies\. Validated experimental neutron flux data using MCNP transport simulations under 14 MeV neutron irradiation conditions\. Optimized High\-Density Polyethylene \(HDPE\) reflector thickness to improve neutron retention and reaction rates within the assembly\.

Summer Intern at Indira Gandhi Centre for Atomic Research \(IGCAR\), Kalpakkam, Jun 2014 \- Jul 2014 · 2 mos, Kalpakkam, India \(near Chennai\), Reviewed advanced materials for nuclear fuel cladding applications and studied fundamentals of fast breeder reactor systems and reactor design\.

## Education

Doctor of Philosophy \- PhD, Computational Materials Science, Université Paris\-Saclay

Master’s Degree, Nuclear Engineering, University of Petroleum and Energy Studies, Dehradun

## Public links

- github: https://github\.com/anshuman513kite

[View the original profile](https://tal.club/meet/anshuman-choudhury)

```json
{
  "@context": "https://schema.org",
  "@graph": [
    {
      "@type": "Person",
      "@id": "https://tal.club/meet/anshuman-choudhury#person",
      "name": "Anshuman Choudhury, PhD",
      "url": "https://tal.club/meet/anshuman-choudhury",
      "image": "https://tal.club/meet/anshuman-choudhury/profile-photo.jpg",
      "description": "Meet Anshuman Choudhury, PhD, Computational Materials Scientist (PhD) | Dislocation Dynamics & MD Simulation | AI/ML for Scientific Computing | Open to Opportunities based in Bangalore Urban. AI stack includes claude, codex, and github copilot. Explore public work links including github.",
      "jobTitle": "Scientific Computing & AI Research",
      "worksFor": {
        "@type": "Organization",
        "name": "Self-Employed"
      },
      "hasOccupation": {
        "@type": "Occupation",
        "name": "Scientific Computing & AI Research"
      },
      "homeLocation": {
        "@type": "Place",
        "name": "Bangalore Urban"
      },
      "alumniOf": [
        {
          "@type": "EducationalOrganization",
          "name": "Université Paris-Saclay"
        },
        {
          "@type": "EducationalOrganization",
          "name": "University of Petroleum and Energy Studies, Dehradun"
        }
      ],
      "knowsAbout": [
        "claude",
        "codex",
        "github copilot",
        "notion",
        "claudecode",
        "ChatGPT"
      ],
      "sameAs": [
        "https://github.com/anshuman513kite"
      ],
      "mainEntityOfPage": {
        "@id": "https://tal.club/meet/anshuman-choudhury#webpage"
      }
    },
    {
      "@type": "ProfilePage",
      "@id": "https://tal.club/meet/anshuman-choudhury#webpage",
      "url": "https://tal.club/meet/anshuman-choudhury",
      "name": "Anshuman Choudhury, PhD on Tal",
      "description": "Meet Anshuman Choudhury, PhD, Computational Materials Scientist (PhD) | Dislocation Dynamics & MD Simulation | AI/ML for Scientific Computing | Open to Opportunities based in Bangalore Urban. AI stack includes claude, codex, and github copilot. Explore public work links including github.",
      "dateModified": "2026-10-08T22:08:57.815Z",
      "isPartOf": {
        "@id": "https://tal.club/#website"
      },
      "mainEntity": {
        "@id": "https://tal.club/meet/anshuman-choudhury#person"
      },
      "primaryImageOfPage": {
        "@type": "ImageObject",
        "url": "https://tal.club/meet/anshuman-choudhury/profile-photo.jpg",
        "contentUrl": "https://tal.club/meet/anshuman-choudhury/profile-photo.jpg",
        "caption": "Anshuman Choudhury, PhD profile photo on Tal"
      },
      "inLanguage": "en-IN"
    }
  ]
}
```