Instagram Audience Persona Analyzer (Compliant)

Classify bios/comments into personas (Student, Tech, Religious, Fitness, Job-seeker, Creator, Business, Other) and generate an aggregate report.
No scraping. Works with official Instagram Graph API data you’re allowed to fetch (audience insights, commenters/likers on your media) or CSV uploads collected with consent.

Analyze your Instagram audience personas Built with Python · FastAPI · Chart.js · CSV Upload · Docker-ready. Just clone → install → run → see charts in your browser 🚀

After following the web api will be running on http://127.0.0.1:8000/ui

git clone https://github.com/perryvegehan/INSTA_AUDIENCE_ANALYSER.git cd INSTA_AUDIENCE_ANALYSER

macOS / Linux python3 -m venv .venv source .venv/bin/activate pip install -r requirements.txt

Windows (PowerShell) python -m venv .venv .venv\Scripts\Activate pip install -r requirements.txt

Run the Demo (no Instagram needed) python -m scripts.demo_classify # Creates data/report.json with audience categories (students, tech, fitness, etc.).

uvicorn app.main:app --reload --port 8000

Open in your browser:

API docs: http://127.0.0.1:8000/docs

Web UI: http://127.0.0.1:8000/ui

👉 In the Web UI:

Click “Load Demo Data” to see sample charts.

Or Upload a CSV with columns id,bio. example csv id,bio user1,B.Tech CSE | Learning DSA and ML user2,Gym Trainer | Calisthenics user3,Entrepreneur | Startup founder | Ecommerce

Customize Personas Edit app/classifier.py → update keywords inside BUCKETS. Restart the server to apply changes.

Run with Docker (optional) docker build -t insta-audience . docker run --rm -p 8000:8000 insta-audience

Then open http://127.0.0.1:8000/ui

Once you are here and understand the code and the working, proceed to add instagram apis and advance the project.

---

TL;DR – Clone & Run

git clone https://github.com/<you>/insta_audience_tool.git
cd insta_audience_tool

make bootstrap     # creates venv + installs deps
make demo          # runs demo on data/sample_bios.csv -> data/report.json
make api           # starts API at http://127.0.0.1:8000/docs

If you don’t have make:

python -m venv .venv
source .venv/bin/activate               # Windows: .venv\Scripts\activate
pip install -r requirements.txt
python -m scripts.demo_classify
uvicorn app.main:app --reload --port 8000

---

Endpoints

• GET /health – liveness probe
• GET /docs/ig – notes on Graph API wiring (no scraping)
• POST /classify – classify items {id,text}
• POST /report – classify + aggregate in one shot

Open Swagger: http://127.0.0.1:8000/docs

Example payload

{
  "items": [
    {"id":"u1","text":"B.Tech CSE @ IIT. Learning DSA and ML. Open to work"},
    {"id":"u2","text":"Gym • Personal Trainer • Calisthenics"},
    {"id":"u3","text":"Entrepreneur | Startup founder | Ecommerce"}
  ]
}

---

Configure Personas

Edit keywords in app/classifier.py under BUCKETS. Tailor for IIT/GATE/tech jobs/creators, etc. Changes take effect on restart.

Optional zero-shot (slower, smarter):

pip install "transformers==4.44.2" torch
export USE_ZEROSHOT=true
export ZS_MODEL=facebook/bart-large-mnli
uvicorn app.main:app --reload

---

Instagram (Compliant Flow)

1. Switch your IG to Business/Creator.
2. Create a Facebook App → get FB_APP_ID / FB_APP_SECRET.
3. OAuth → Page token → exchange for long‑lived Page token.
4. Get IG Business Account ID.
5. Pull audience insights and comments/likers (where permitted).
6. Feed comment/bio text (where consented) into POST /classify; aggregate via POST /report.

Followers’ bios are not exposed via Graph API. Don’t scrape. See app/ig_api.py and GET /docs/ig for endpoint notes.

---

Docker

docker build -t insta-audience .
docker run --rm -p 8000:8000 --env-file .env.example insta-audience
# http://127.0.0.1:8000/docs

Compose:

docker-compose up --build

---

Devcontainer (VS Code)

Open folder → “Reopen in Container”. You’ll get Python with dependencies preinstalled.

---

Contributing & Troubleshooting

• See CONTRIBUTING.md for PR guidelines.
• See TROUBLESHOOTING.md for fixes like ModuleNotFoundError: app, uvicorn not found, and Graph API 403s.

---

License

MIT. See LICENSE.