Writing emails every day is repetitive, time-consuming, and mentally exhausting. This project is an AI-powered email assistant designed to reduce cognitive effort, speed up communication, and improve clarity by generating context-aware emails and concise summaries.
The system runs entirely on a local language model, making it fast, private, and practical for daily use.
Modern communication involves reading and writing large volumes of email. This leads to:
- Repeated drafting of similar emails
- Difficulty understanding long messages quickly
- Increased cognitive load during communication-heavy workflows
- Privacy concerns with cloud-based AI tools
This project addresses these problems by automating email writing and simplifying long content, reducing manual drafting effort by approximately 60 percent.
The assistant generates complete emails by understanding:
- User intent
- Message context
- Desired writing tone
Supported tones include:
- Formal
- Professional
- Friendly
- Polite
- Concise
Any email or long text can be instantly transformed into:
- Tweet-length summaries
- One-line summaries
- TL;DR versions
This makes it easier to quickly grasp long messages without reading everything.
- Runs completely offline
- No external APIs
- No data leaves the system
The assistant uses a local Llama 3.1 model through Ollama, ensuring full data privacy.
User Input ↓ Prompt Engineering Layer ↓ LangChain4j ↓ Local LLM (Llama 3.1 via Ollama) ↓ Generated Email or Summary
- Programming Language: Java
- LLM Framework: LangChain4j
- Language Model: Llama 3.1 running locally via Ollama
- Prompt Design: Context-aware and tone-controlled prompts
- Execution Mode: Fully local
- Reduced manual email drafting effort by approximately 60 percent
- Improved clarity and readability of long emails
- Faster understanding of complex or lengthy messages
- Zero dependency on cloud-based AI services
- Java version 17 or higher
- Ollama installed locally
ollama pull llama3.1
./gradlew run
- Writing professional emails quickly
- Summarizing long email threads
- Generating short, clear replies
- Reducing mental fatigue from constant communication
- Email thread summarization
- Grammar and tone refinement suggestions
- Gmail and Outlook integration
- Web-based user interface
Developed an AI-powered email assistant using LangChain4j and a local Llama 3.1 model via Ollama to generate context-awa