Built by
Who's behind VibeAlchemy.
MJ
Manu Joseph V
Principal Data Scientist
Bengaluru, India
LinkedIn manujosephv GitHub manujosephv X / Twitter @manujosephv Email manujosephv@gmail.com

Manu is a self-taught data scientist with 15+ years of experience working with Fortune 500 companies on machine learning and AI at scale. He has deep expertise in time series forecasting, deep learning for tabular data, and — more recently — large language models and agentic systems.

He built VibeAlchemy because he kept watching good ideas die in planning docs. He knew exactly how to build the AI infrastructure to fix that. So he did.

1.6k
GitHub stars on
PyTorch Tabular
4
arXiv research
papers published
2nd
Edition of his
time series book
Creator of PyTorch Tabular — an open-source framework for deep learning with tabular data. Over 1,600 GitHub stars, thousands of monthly downloads, cited in academic research.
Author of Modern Time Series Forecasting with Python (Packt, 2nd ed.) — a practitioner reference used by data scientists at Amazon, Microsoft, and Kraft Heinz.
Published researcher on deep learning architectures for tabular data — including GATE and GANDALF, available on arXiv. Speaker at AI/ML conferences.
The same rigour that goes into building production ML systems at scale went into building this pipeline. VibeAlchemy is not a wrapper around a chatbot — it's a deliberate, multi-stage system designed to think before it writes.