Skip to content

Folders and files

NameName
Last commit message
Last commit date

Latest commit

Β 

History

3 Commits
Β 
Β 

Repository files navigation

Hi, I'm Sreekumar

Scientist turned AI builder working at the intersection of machine learning, scientific computing, and real-world systems.

I come from a background in astrophysics and Bayesian inference, where I developed high-performance computational methods to analyse cosmological data from space telescopes. My research contributed to the discovery of three exoplanets through large-scale C++ simulations.

Today I focus on building robust, efficient AI systems that translate scientific ideas into practical tools. My work has ranged from Bayesian deep learning for financial forecasting to multi-modal machine learning platforms for real estate analytics and underwriting.

I enjoy working with small, ambitious teams exploring new applications of AI and building systems from first principles.


Current Exploration

⚑ Energy-efficient machine learning
🧠 Scientific machine learning and probabilistic models
🌍 AI for geospatial intelligence and remote sensing
⛏️ Mineral discovery using ML and satellite data
πŸ” Efficient adaptations of foundation models


Technical Stack

Languages

  • Python
  • C++

Machine Learning

  • PyTorch
  • Scikit-learn
  • Probabilistic modelling

Scientific Computing

  • High Performance Computing (HPC)
  • Bayesian inference
  • Numerical simulation

Domains

  • Geospatial & remote sensing
  • Financial modelling
  • Real estate market intelligence

Selected Experience

Machine Learning Leadership β€” Real Estate AI
Led development of a multi-modal ML platform combining tabular, text, and imagery data to forecast property markets and support underwriting decisions.

Entrepreneurship β€” AI in Finance
Co-founded a machine learning hedge fund through Entrepreneur First, developing Bayesian deep learning models for financial forecasting.

Scientific Research β€” Astrophysics
Developed Bayesian statistical methods and HPC software for analysing cosmological datasets from major space telescope missions.


Fun Fact

πŸ›°οΈ I once co-wrote a C++ simulation to model exoplanet orbits β€” and the work contributed to the discovery of three planets.


Let's Connect

I'm always interested in discussing ideas at the intersection of:

  • machine learning
  • science-inspired AI
  • geospatial intelligence
  • efficient and data-lean AI systems

If you're building something interesting in this space, I'd love to hear about it.

About

My personal repository.

Resources

Stars

Watchers

Forks

Releases

Packages

Contributors