Kanishka Fartiyal

Claims you can check

Each card points at something real — open the source and verify for yourself.

• check_circle_outlineStateddescriptionRésumé

  E-Commerce Sales & Customer Segmentation

  Cleaned and analysed 96,477 real orders from the Olist dataset across 9 product categories and 3 years of transaction data to uncover revenue and behaviour trends. Applied statistical clustering (KMeans) to segment customers into 4 actionable groups (Champions, At-Risk, Hibernating, Lost) with a Silhouette Score of 0.525; found the top 20% of customers drove ~62% of revenue — a pricing/retention-relevant insight.
• check_circle_outlineStateddescriptionRésumé

  Tech Industry Layoffs & Data Science Salary Analysis

  Cleaned and standardised 746,809 records across 1,703 companies (2020–2025) using advanced SQL (CTEs, window functions, CASE WHEN) to answer 8 business questions on trends, geography and funding. Cross-referenced with a salaries dataset and identified 14 companies that shut down despite raising substantial funding, with activity spiking after the ChatGPT launch — visualised across 12 queries with Seaborn.
• check_circle_outlineStateddescriptionRésumé

  Data Analyst Intern @ Ingrify (Verinova Labs)

  Developed a process to collect, validate and clean 499 real-world product records across brands, categories and nutrition — standardising messy source data into a decision-ready dashboard and cutting manual lookup time by ~40%. Ran data validation and quality checks across categories while building 12+ FastAPI/SQLAlchemy/MySQL endpoints to power search, filtering, and reporting for business users on consistently validated data. Analysed attributes across 6 classifications and 30+ brand categories, translating validated data into visual comparisons used for business reporting and competitive benchmarking.
• check_circle_outlineStateddescriptionRésumé

  Content Analyst @ Content Analytics (Freelance)

  Tracked views, engagement and retention metrics across client deliverables in Excel, identifying performance trends and building repeatable reporting templates. Ran an informal test comparing content formats — found hook delivery under 3 seconds plus mid-video pattern interrupts drove 2× average watch-time — turned the finding into a documented, repeatable brief for clients.
• check_circle_outlineStateddescriptionRésumé

  Excel Dashboards for Business Analytics

  Certification by Great Learning
• check_circle_outlineStateddescriptionRésumé

  Data Analysis using Excel

  Certification by Great Learning
• check_circle_outlineStateddescriptionRésumé

  CS50P – Introduction to Programming with Python

  Certification by Harvard / edX
• check_circle_outlineStateddescriptionRésumé

  Data Analysis – Beginner to Advanced

  Certification by Udemy