https://github.com/shivendrra/seeker
Research Application based on AI Agentic workflow
https://github.com/shivendrra/seeker
ai artificial-intelligence finetuning google-custom-search-api google-custom-search-engine llama2 llm numpy pandas python pytorch react-native reactjs transformers youtube-api
Last synced: 2 months ago
JSON representation
Research Application based on AI Agentic workflow
- Host: GitHub
- URL: https://github.com/shivendrra/seeker
- Owner: shivendrra
- Created: 2024-01-01T08:38:18.000Z (over 2 years ago)
- Default Branch: main
- Last Pushed: 2024-01-27T19:07:08.000Z (over 2 years ago)
- Last Synced: 2025-08-28T20:38:13.908Z (10 months ago)
- Topics: ai, artificial-intelligence, finetuning, google-custom-search-api, google-custom-search-engine, llama2, llm, numpy, pandas, python, pytorch, react-native, reactjs, transformers, youtube-api
- Language: Python
- Homepage:
- Size: 35.8 MB
- Stars: 2
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Seeker: AI Agent-based Research Application
An autonomous AI agent to assist journalists, lawyers, and academic researchers with complex research tasks, providing structured answers with clear citations and provenance.
## Introduction
In professions like journalism, law, and academic research, the ability to quickly gather, synthesize, and cite information from vast sources is critical. Traditional search methods are often time-consuming and linear. **Seeker** is an advanced research assistant powered by the Google Gemini API, designed to automate and enhance this process.
Seeker acts as an autonomous agent that interprets complex user queries, plans a multi-step research workflow, utilizes various tools (like document and web search), and synthesizes the findings into a structured, well-cited response. It's built to be a reliable partner for professionals who demand accuracy, transparency, and efficiency in their research.
## Key Features
- **Autonomous AI Agent**: Interprets natural language requests and plans a multi-step research strategy.
- **Cited & Verifiable Answers**: Provides answers with clear citations, including source, page/paragraph, and date, to ensure provenance and trust.
- **Transparent Process**: Features a **Trace View** that shows the AI's exact plan and execution steps, allowing for full auditability of the research process.
- **Long-Term Memory**: Remembers context across sessions to personalize the experience and build upon previous research.
- **Multi-Tool Capability**: Designed to use a suite of tools, including internal document retrieval and external web searches, to gather comprehensive information.
- **Secure & Private**: Built on Firebase for secure user authentication and data storage.
- **Modern UI/UX**: A clean, responsive, and intuitive interface with both **Light and Dark modes**, designed for focus and productivity.
- **Session Management**: Organizes research into distinct, titled sessions that can be revisited or deleted.
## Tech Stack
- **Frontend**: [React](https://reactjs.org/), [TypeScript](https://www.typescriptlang.org/), [Tailwind CSS](https://tailwindcss.com/)
- **Backend & Database**: [Firebase](https://firebase.google.com/) (Authentication, Firestore for data storage)
- **AI Model**: [Google Gemini API](https://ai.google.dev/) (using `gemini-2.5-pro`)
- **Development/Build**: [Vite](https://vitejs.dev/)
## Project Structure
The project is organized into a standard React application structure:
```
/
├── app/
│ ├── components/ # Reusable React components
│ ├── services/ # Modules for external services (Firebase, Gemini)
│ ├── hooks/ # Custom React hooks (e.g., useAuth)
│ ├── utils/ # Helper functions (e.g., response parsing)
│ ├── App.tsx # Main application component
│ ├── index.tsx # Entry point of the React app
│ ├── types.ts # TypeScript type definitions
│ └── constants.ts # Core constants, including the main system prompt
├── .gitignore
└── package.json
```
vibe coded using GoogleAI Studio