https://github.com/akhilk2802/helix
Agentic Recruiter
https://github.com/akhilk2802/helix
agentic-ai flask llm openai-api react
Last synced: about 2 months ago
JSON representation
Agentic Recruiter
- Host: GitHub
- URL: https://github.com/akhilk2802/helix
- Owner: akhilk2802
- Created: 2025-05-23T00:47:30.000Z (about 1 year ago)
- Default Branch: main
- Last Pushed: 2025-06-05T14:34:08.000Z (about 1 year ago)
- Last Synced: 2025-06-08T20:37:17.333Z (about 1 year ago)
- Topics: agentic-ai, flask, llm, openai-api, react
- Language: Python
- Homepage:
- Size: 12.3 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Helix - The Agentic Recruiter
Helix is a real-time AI-powered recruiter assistant designed to streamline the outreach process for recruiters. With Helix, recruiters can chat naturally, generate and edit personalized multi-step outreach sequences, and manage sequences manually through a workspace UI.
This tool is built with a combination of **React**, **Flask**, **PostgreSQL**, **Socket.IO**, **OpenAI GPT-4**, and **ChromaDB** to provide real-time interaction, memory recall, and intelligent sequence creation.
---
## Features
- **Conversational AI** – Recruiters can talk to Helix naturally to create or modify outreach messages.
- **Sequence Editor** – AI-generated recruiting sequences are editable directly in the UI with autosave support.
- **Tool Call Detection** – Detects recruiter intent and dynamically calls backend tools (`create_sequence`, `edit_sequence`, `delete_step`).
- **Workspace View** – Visually structured sequence steps for LinkedIn and Email messaging.
- **Memory Integration** – Uses ChromaDB to remember recruiter-specific messages and context for future use.
- **Real-time Updates** – Powered by Socket.IO for low-latency interactions.
---
## Tech Stack
| Layer | Technology |
|--------------|----------------------------------|
| Frontend | React, TypeScript, Bootstrap |
| Backend | Python, Flask, Flask-SocketIO |
| Database | PostgreSQL (relational DB) |
| Vector Store | ChromaDB |
| Embeddings | OpenAI Embedding API |
| LLM | OpenAI GPT-4 |
| Messaging | Socket.IO (real-time messaging) |
---
## Setting up locally
### 1. Unzip the project
```bash
cd helix
```
### 2. Backend Setup - Flask
- Prerequisites
- Python 3.0 +
- PostgreSQL
- OpenAI API Key
- ChromaDB
- create and setup environment variables in ```.env``` file
- ### Example .env file
```bash
DATABASE_URL=
CHROMA_DB_PORT=
OPENAI_API_KEY=
CHROMA_DB_URL=
```
```bash
cd backend
python3 -n venv .env
source venv/bin/activate
```
- Install dependencies
```bash
pip install -r requirements.txt
```
- Start ChromaDB (as HTTP Server)
```bash
chroma run --host localhost --port 8000
```
- Run the backend server
```bash
flask run
```
### 3. Frontend Setup - React
- Prerequisites
- node 18+
- npm
- Create and Setup ```.env``` file
- Example ```.env``` file
```
REACT_APP_API_URL=
REACT_APP_PORT=
```
- Start the frontend
```bash
npm start
```
**Usually starts React at https://localhost:3000, unless a specific port is specified.**
## Example workflow
1. Log in with your name and company (mocked for now).
2. Chat with Helix: “Create a sequence for a Backend Engineer in New York with 5 years experience.”
3. Watch Helix respond with a recruiting sequence.
4. Click into the workspace panel and:
- Manually edit steps with autosave.
- Delete any step using the 🗑 icon.
- Ask Helix to edit or append new steps via chat.
## How the UI looks
### Login and Signup page
### Home page (after login)
### Architecture of the MVP
