Skip to content
View kairav11's full-sized avatar

Block or report kairav11

Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
kairav11/README.md

πŸ’« About Me

Hi, i’m Kairav Sanghvi πŸ‘‹
Data Science graduate focused on product analytics, machine learning, and growth experimentation.
Based in Berlin, Germany.

Good questions matter more than perfect answers.


About Me

I work at the intersection of data, product, and growth.

My work is centered around:

  • analytics dashboards that make metrics easy to act on
  • automated workflows that reduce manual effort
  • data pipelines that support product and marketing decisions

I enjoy taking messy, real-world data and turning it into insights teams can actually use.


What I Work On

  • Product & Growth Analytics
    Funnel analysis, activation and retention tracking, KPI monitoring

  • Data Analysis & Experimentation
    EDA, hypothesis testing, A/B analysis, statistical reasoning

  • Machine Learning
    Regression, classification, clustering, feature engineering

  • Automation & LLMs
    Workflow automation with n8n and Zapier, GPT-based feedback and content analysis


Tech Stack

Languages

  • Python
  • SQL

Data & Analytics

  • Pandas, NumPy
  • Statistical analysis, KPI tracking, A/B testing

Machine Learning

  • Scikit-learn
  • Regression, classification, clustering

Visualization & BI

  • Metabase
  • Power BI
  • Matplotlib, Seaborn

Automation & Tools

  • n8n, Zapier
  • Git, GitHub
  • HubSpot
  • PostgreSQL

Python MySQL MicrosoftSQLServer SQLite Postgres C C++ Java R Go JavaScript TypeScript AWS Pandas Supabase Adobe Canva Figma Matplotlib NumPy Pandas Plotly PyTorch scikit-learn Scipy TensorFlow GitHub Git Notion Power Bi


Featured Project

ETA Prediction and Last-Mile Delivery Optimization

πŸ”— Repository

Built and evaluated ETA prediction models using real-world logistics data.

Highlights

  • Extensive EDA and feature engineering, including GPS and timestamp correction
  • Compared regression and ensemble models (Random Forest, XGBoost, CatBoost)
  • Evaluated performance using MAE, RMSE, RΒ², and tolerance-based accuracy
  • Integrated spatio-temporal context such as traffic and weather
  • Found ensemble models to outperform deep learning in sparse-data settings

Current Focus

  • Building clearer, more actionable BI dashboards
  • Improving workflow automation with n8n and CRM tools
  • Applying growth analytics frameworks to product-led businesses
  • Exploring data applications in open-source and cybersecurity ecosystems

Let’s Connect

πŸ“§ Email
πŸ”— LinkedIn

🌐 Socials:

LinkedIn email

πŸ“Š GitHub Stats:




Popular repositories Loading

  1. Customer_Rating_Prediction Customer_Rating_Prediction Public

    Jupyter Notebook

  2. kairav11.github.io kairav11.github.io Public

    Portfolio Website

  3. kairav11 kairav11 Public

  4. Last_Mile_Delivery_Analysis_and_Prediction Last_Mile_Delivery_Analysis_and_Prediction Public

    Jupyter Notebook

  5. Bike_Share_Demand_Analysis Bike_Share_Demand_Analysis Public

  6. Fine-Tuning-an-LLM-using-mini-platypus-Dataset Fine-Tuning-an-LLM-using-mini-platypus-Dataset Public

    This project demonstrates the process of fine-tuning a pre-trained Large Language Model (LLM) using QLoRA, and provides a simple interface to compare model responses before and after fine-tuning. I…

    Jupyter Notebook