Skip to content

A containerised microservice powered by FastAPI, LangChain, and Gemini 1.5 that acts as a general-purpose AI assistant. It can perform calculations, summarise text, parse CSVs and calendar files, simulate booking meetings, and search the web — making it an ideal backend for productivity tools, Slack bots, or internal assistants.

Notifications You must be signed in to change notification settings

Peippo1/Focus-Agent

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Python License FastAPI Dockerized

🧠 Focus Agent — Productivity AI Assistant

focus-agent is a containerized microservice built with FastAPI, LangChain, and Gemini 1.5. It acts as a general-purpose AI productivity assistant that can handle a variety of tasks including calculations, summarisation, file parsing, scheduling mock meetings, and web search.

✨ Features

  • ✅ REST API via FastAPI
  • 🧠 LLM agent powered by Gemini 1.5
  • 🔌 Modular tools:
    • Calculator
    • Summariser
    • CSV parser
    • Mock meeting scheduler
    • Web search
  • 🐳 Dockerized for deployment
  • 🛠 DevOps-ready: .env, Makefile

📦 Installation

git clone https://github.com/Peippo1/Focus-Agent.git
cd focus-agent
python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt

🚀 Running the App

make run

Or with Docker:

make docker-build
make docker-run

🔧 API Endpoints

  • GET /ping — Health check
  • POST /agent — Query the agent
    • Body: { "query": "Summarise this text", "file": <optional file> }

📝 License

MIT License


Built by @Peippo1

About

A containerised microservice powered by FastAPI, LangChain, and Gemini 1.5 that acts as a general-purpose AI assistant. It can perform calculations, summarise text, parse CSVs and calendar files, simulate booking meetings, and search the web — making it an ideal backend for productivity tools, Slack bots, or internal assistants.

Topics

Resources

Stars

Watchers

Forks

Packages

No packages published