A small FastAPI app for running People Data Labs person searches, enriching matches with AI, and stepping through LinkedIn and email drafts before you send anything. Built for people that NEED their next internship and want to take THEIR future into their own hands.

Internflow.Demo.mp4

---

Search → browse → configure in one place.

Search Filters, dataset, run query	Browse Grid, filters, export to configure
Lead detail Modal on Browse — full profile	Configure Sender profile, drafts, queue

---

THIS IS NOT A PRODUCTION PROJECT DEPLOYED TO A WEBSITE AS OF NOW. To use simply clone and follow the steps indicated below.

---

When to use it

You need…	Intern gives you
Filtered PDL searches with scroll tokens and credit-aware batching	Search UI with industry, location, dataset, and optional company filters
A queue of new leads while AI scores and reasons on each profile	Browse with filters, selection, and export into a configure session
Ready-to-edit outreach tied to your sender name and portfolio link	Configure: one lead at a time, copy buttons, Done / Deleted to clear the queue

---

How it flows

1. Search — Run a PDL query (size capped, same filters reuse scroll state where applicable). Pending rows land in the database while enrichment runs.
2. Browse — Filter, open detail, select leads, Export to configure.
3. Configure — Enter sender details once, then walk the stack: copy drafts, mark Done or swipe/delete, until the export is finished.

---

Stack

• Python 3.10+, FastAPI, Jinja2, SQLModel / SQLite (config.yaml points at the DB path)
• Tailwind CSS built locally (npm run build:css); styles live in templates/ and ship as static/css/app.css
• OpenAI (or your configured provider) for scoring and drafts in the enrichment path
• People Data Labs API for search and enrichment (API key from environment)

---

Quick start

git clone <your-repo-url> internflow
cd internflow
python -m venv .venv
.venv\Scripts\activate          # Windows
# source .venv/bin/activate       # macOS / Linux

pip install -r requirements.txt

Set PDL_API_KEY and OPENAI_API_KEY (if you use enrichment). The app loads them from the environment (python-dotenv reads a .env file in the project root if present).

Build CSS once (or whenever templates change):

npm install
npm run build:css

Either run it via python or use your VSCode inbuilt run and debug functionality.

Default bind: http://127.0.0.1:5000 (see main.py).

---

Configuration

File / place	Role
config.yaml	SQLite path and DB flags
Environment / .env	API keys

---

Contributing

Issues and PRs welcome. After changing HTML classes under templates/, run npm run build:css but don't commit your static/css/app.css.