-->

H

e

y

 

T

h

e

r

e

,

 

I

'

m

J

a

y

 

S

i

n

g

h

.

I

u

s

e

d

a

t

a

t

o

d

r

i

v

e

i

m

p

a

c

t

I’m a data enthusiast and code whisperer who enjoys making machines a little smarter (and occasionally catching them hallucinating). I work at the intersection of models, math, and messy datasets — where bugs live and insights hide. Whether it’s tweaking prompts, training models, or building pipelines that don’t crash on Monday mornings, I’m all in.

I love the thrill of solving weird problems, breaking things just to fix them better, and making code that’s not just functional, but kind of fun to write. Always learning, always shipping — preferably before the coffee runs out 💻.

   

AboutMe

I am an IIT Kharagpur graduate. I have been driven by a profound passion for harnessing advanced machine learning to solve real-world challenges for the past 4 years. My journey has been enriched by impactful work experiences. I worked as a Data Scientist at SymX.ai, worked as Data Science Intern at Innovaccer, Affine, Cambridge Judge Business School, and others.


My technical foundation spans core machine learning, data analysis, deep learning, large language models, and Python web frameworks which I reflect on Kaggle profile. Being a regular open-source contributor, I try to grow, adap, learn from this ever growing developer community. My enthusiasm lies in constructing sophisticated models like recommendation systems, time series, generative, and predictive models, particularly for projects that yield impactful results, find those on my github.


My drive extends beyond the intellectual rigors of my field; I find that true peak performance comes from a balanced life. The adrenaline of high-intensity sports, the discipline of sculpting a strong physique, and the tranquility found in spiritual exploration are vital counterpoints that sharpen my mind and enrich my approach to everything I do.

Jay Singh

Skills

Tech Stack

  • PYTHON
  • TENSORFLOW
  • PYTORCH
  • LANGCHAIN
  • AWS
  • HUGGINGFACE (TRANSFORMER MODELS)
  • DBMS
  • FLASK
  • DJANGO
  • MLFLOW
  • DOCKER
  • KUBERNETES
  • N8N
  • MCP Protocol
  • GITHUB
  • GIT
  • HTML
  • CSS

Projects

pexelicon website favicon
RNN / LSTM Foundations

See how RNN and LSTM functions under the hood. Written in plain python code.

Flow image
pexelicon website favicon
Transformer-Model-From-Scratch

See the replication of "Attention is all you need" paper.

Architecture Overview
pexelicon website favicon
Semantic and LLM based Search/Reranker

Semantic and LLM based retrieval/reranking comparison with other traditional search indexers/rankers.

Architecture Overview
pexelicon website favicon
Reinforcement Learning with Human Feedback Prototype

A small prototype of reward model and its working in RLHF on Reward Bench dataset.

Architecture Overview
pexelicon website favicon
Multi Agentic Interviewer

A multi agentic resume parser+interviewer for realtime technical interviews.

Architecture Overview
pexelicon website favicon
Gmail Draft Duplication Model

A Flask based web app that uses Gmail client to duplicate Gmail on mili second latency.

Architecture Overview