Speculative systems and visual studies
A growing collection of work about emerging technology and human experience.
Research engineer · Computational physicist
AnshThukral
Building intelligent systems from first principles—where rigorous experiments meet useful products.
Explore selected work ↓01 / Approach
A way of working
I’m interested in the space between understanding a system and making it work.
My background in computational physics shapes how I build: isolate assumptions, design the experiment, follow the evidence, then turn what survives into something robust.
02 / Selected work
Case studies in rigor.
Research and systems spanning gravitational waves, financial forecasting, quantum simulation, and particle physics.
LIGO / DeepClean
A simulation-led investigation of deep learning for environmental noise subtraction in gravitational-wave strain data.
- PyTorch
- Signal processing
- Time series
Financial ML
A production-minded forecasting and portfolio optimization system built for reliable daily inference.
- Forecasting
- MLOps
- Optimization
QPT / TFIM
A numerical study connecting quantum phase transitions to a classical two-dimensional Ising mapping.
- Simulation
- Numerical methods
- Quantum
Higgs classification
Controlled benchmarking across neural and tree-based models on simulated ATLAS collision events.
- Benchmarking
- PyTorch
- Kubernetes
03 / In progress
ActiveResearchFlow
An AI-native workspace for the way research actually unfolds.
Currently exploring how knowledge, evidence, and agentic workflows can coexist without flattening the ambiguity of real research.
04 / Garden
Notes from the edges.
Design studies, unfinished experiments, writing, and the ideas currently occupying my attention.
Research notes, slowly made legible
Working through machine learning, physics, and products by writing about them.
Learning in public, selectively.
Current questions, tools, and threads worth returning to.
Portrait
coming into focus
coming into focus
05 / About
Physics taught me to question the model. Design taught me to question the frame.
I’m a computational physicist by training and a research engineer by practice. I care about technically honest work, well-designed tools, and explanations that respect complexity without hiding behind it.
More about me ↗06 / Contact
Open to research-engineering conversations