Blyton/Netflix-content-analysis-dashboard

🎬 Netflix Content Analysis Dashboard

A premium, Netflix-themed analytics platform built with Streamlit — not just a chart page, but a full product...

🎬 Netflix Content Analysis Dashboard

A premium, Netflix-themed analytics platform built with Streamlit — not just a chart page, but a full product: hero landing, KPI dashboard, interactive Plotly visualizations, a live SQL query layer, a Python cleaning notebook, a rule-based insight generator, a world map, a storytelling page, and a Power BI gallery.

Quickstart

pip install -r requirements.txt
streamlit run app.py

The app opens at http://localhost:8501.

Project structure

netflix_dashboard/
├── app.py                     # entry point — nav, sidebar filters, page routing
├── data/
│   ├── netflix_titles_raw.csv     # your uploaded dataset, untouched
│   └── netflix_titles_clean.csv   # cleaned version the app reads (see modules/sections/python_section.py for the pipeline)
├── modules/
│   ├── data_loader.py         # cached loading + sidebar-filter logic
│   ├── styles.py               # design tokens, CSS injection, grain background, KPI cards
│   ├── components.py           # animated counters, movie card HTML
│   └── sections/
│       ├── hero.py             # landing page
│       ├── dashboard.py        # KPIs + 9 Plotly charts
│       ├── explorer.py         # search + movie card grid
│       ├── worldmap.py         # choropleth + spotlighted markets
│       ├── insights.py         # rule-based AI Insight Generator
│       ├── sql_section.py      # live SQLite query cards + custom query box
│       ├── python_section.py   # cleaning code, missing-value charts, live pandas snippets
│       ├── powerbi_section.py  # screenshot gallery (auto-loads from assets/powerbi/)
│       ├── story.py             # data-driven narrative page
│       └── about.py             # profile, skills, links, resume
├── assets/
│   ├── powerbi/                # drop your Power BI screenshots here (png/jpg)
│   └── resume.pdf              # add your resume here to enable the download button
└── requirements.txt

Notes on scope

• —Posters: the Movie Explorer uses styled placeholder cards (initials on a gradient tile) rather than fetched poster art, by design — no external image API is wired in.
• —AI Insights: fully rule-based. Every line is computed live from the filtered dataframe (percentages, leading genre/country, peak years, etc.) — no API key required, works offline.
• —Power BI: the gallery auto-populates from assets/powerbi/. Export your dashboard pages as images and drop them in; no screenshots ship with this repo.
• —SQL: queries run against a real in-memory SQLite database built from your data on each session — nothing is hardcoded, including the custom query box at the bottom of that page.

Data cleaning summary

Raw dataset: 8,807 rows. After filling missing director/cast/country with "Unknown", dropping the handful of rows missing rating/duration, and de-duplicating on (title, type, release_year): 8,800 rows — 6,126 Movies / 2,674 TV Shows. Full pipeline is in modules/sections/python_section.py (Cleaning Code tab) and mirrored below.