ROHAN PATNAIK

DATA SCIENTIST

GitHub Light GitHub Dark LinkedIn Light LinkedIn Dark Medium Light Medium Dark Kaggle Light Kaggle Dark

I'm a data scientist with hands-on experience building machine learning systems and predictive modeling using Python, SQL & stats!
I've honed my skills in financial data classification, fraud detection and Natural Language Processing.

Resume
Profile picture

PROJECTS

Fake news identifer

Classifies news as true or untrue using a bidirectional LSTM model.

AI article writer

Generate tailored articles by specifying a topic and word count limit.

Decensored DeepseekR1

A VS Code extension for an AI-powered chat assistant within VSCode.

Is This Food

A web app to detect food in images using a TensorFlow Lite model.

ARTICLES

Time Series Analysis Article

A comprehensive Time Series Analysis roadmap

A one-stop resource for learning and practicing time series analysis (TSA) from start to finish.

 Hugging Face GGUF Model Article

Run any huggingface GGUF model locally on windows

In this post I’ll walk you through the process of downloading a GGUF model from Hugging Face and running it locally with Ollama.

 Hugging Face GGUF Model Article

From YouTube to Answers: Building a RAG System on youtube transcripts

In this blog post, we’ll explore how to transform a YouTube video into a powerful personal Q&A engine!

SKILLS

Pytorch, Keras, Scikit-Learn, TensorFlow, A/B testing, Excel, Tableau, NLP, Kubeflow Pipelines, Timeseries

MySQL, PostgreSQL, Trino, MinIO

Python, Flask, Jenkins, TypeScript, Docker, Kubernetes, Git, GitHub Actions

WORK EXPERIENCE

Senior Analyst

Tiger Analytics
Dec 2024 – Present
  • Worked on the McDonalds Insights project by analyzing sales and revenue generation queries using SQL and Python.
  • Developed dashboards to provide actionable insights for business strategy.

Data Scientist

National Payment Corporation of India (NPCI)
April 2024 - October 2024
  • Led the development of fraud detection for AePS by engineering raw data features to create predictive models—reducing fraud incidents by 1.2 million while processing over 4TB of data.
  • Utilized Python to implement supervised machine learning techniques for time series anomaly detection.
  • Contributed to a pioneering UPI project by developing a federated model for financial fraud detection, handling complex real-world data for new onboarding banks.

Data Scientist / Cofounder

SKIPCHECKR AI
  • Spearheaded development of an innovative self-checkout system powered by machine learning, from concept to MVP.
  • Demonstrated strong leadership and project management by driving sub-projects to completion.
  • Applied data science and computer vision in retail to enhance efficiency through targeted research and A/B testing.

Data Scientist / Software Engineer

SIEMENS
July 2019 – April 2022
  • Developed and analyzed sound and numeric data for client assets using computer vision, achieving 87% accuracy in fault detection via RUL classification.
  • Designed dashboard views with API calls to monitor asset health while performing ML analysis.
  • Led team communications to address key security issues arising from tokenization in the user app.

CONTACT ME!