Portfolio Assistant

A sophisticated LangGraph-powered AI assistant designed to showcase Robin Rathore's development team and services. This assistant uses LangChain, Google's Gemini model, and a state-based workflow to handle portfolio queries, generate project proposals, and communicate with potential clients.

Overview

This project implements an AI portfolio assistant with the following capabilities:

• Share information about team projects and members
• Generate professional project proposals
• Provide details about services offered
• Send emails with project information
• Maintain proper security constraints

Table of Contents

• Features
• Prerequisites
• Installation
• Environment Variables
• Project Structure
• Core Components
• Tools
• Workflows
• Security
• Usage Examples
• Contributing

Features

• Project Information: Access details about team projects, including technologies used and team composition
• Team Member Profiles: Get information about individual team members and their skills
• Service Information: Explore the services offered by Robin's development team
• Project Proposals: Generate professional proposals based on client requirements
• Secure Email Communication: Send emails directly to Robin with proper security measures
• State-based Workflow: Advanced LangGraph implementation for controlled conversations

Prerequisites

• Node.js (v16.x or higher)
• npm or yarn
• Google Gemini API key
• Gmail account for sending emails

Installation

1. Clone the repository:

git clone https://github.com/robinrathore/portfolio-assistant.git
cd portfolio-assistant

2. Install dependencies:

npm install

3. Create a .env file in the project root (see Environment Variables section)

4. Run the application:

npm start

Environment Variables

Create a .env file with the following variables:

# Google Gemini API
GOOGLE_API_KEY=your_google_api_key_here

# Email configuration
EMAIL=your_gmail_address
PASSWORD=your_app_specific_password

Note: For Gmail, you'll need to use an App Password instead of your regular password. See Google's documentation for instructions.

Project Structure

The main functionality is implemented in agent.ts, which contains:

• LLM configuration
• Tool definitions
• State annotations
• Graph workflows
• Helper functions

Core Components

LLM Configuration

The assistant uses Google's Gemini 1.5 Pro model:

const llm = new ChatGoogleGenerativeAI({
  model: "gemini-1.5-pro",
  temperature: 0,
  maxRetries: 2,
});

State Annotations

A custom state annotation for managing portfolio workflows:

const PortfolioStateAnnotation = Annotation.Root({
  request: Annotation,
  response: Annotation,
  status: Annotation,
  feedback: Annotation,
  projectDetails: Annotation,
});

Schemas

Zod schemas define structured data for:

• Project information
• Email content
• Project proposals

Tools

1. sendEmail

Sends emails with project information to a fixed recipient for security.

Features:

• Restricted to a single recipient (robinsingh248142@gmail.com)
• Prevents duplicate emails in a session
• Supports priority levels

Usage:

await sendEmail({
  to: "robinsingh248142@gmail.com",
  subject: "Project Proposal",
  body: "<h1>Project Proposal</h1><p>Details here...</p>",
  priority: "normal"
});

2. getProjectInfo

Retrieves detailed information about specific projects.

Projects available:

• Website Redesign
• Mobile App
• Data Dashboard
• E-commerce Platform
• CRM System

Usage:

await getProjectInfo({ projectName: "Website Redesign" });

3. getTeamMemberInfo

Provides information about team members, their skills, and projects.

Team members available:

• Robin (Lead)
• Alex, Jamie, Morgan, Taylor, Casey, Jordan (Team members)

Usage:

await getTeamMemberInfo({ memberName: "Robin" });

4. generateProjectProposal

Creates a tailored project proposal based on client requirements.

Customization options:

• Client name
• Project type
• Requirements
• Budget
• Timeline

Usage:

await generateProjectProposal({
  clientName: "TechCorp",
  projectType: "Web Application",
  requirements: "E-commerce platform with payment processing",
  budget: "$10,000-$15,000",
  timeline: "3 months"
});

5. introduceTeam

Provides a comprehensive introduction to Robin's development team.

Detail levels:

• brief: Basic team information
• detailed: Complete team profile with testimonials

Usage:

await introduceTeam({ detail: "detailed" });

6. getTeamServices

Retrieves information about services offered by the team.

Services covered:

• Web Development
• Mobile Development
• UI/UX Design
• Custom Software Development
• E-commerce Solutions

Usage:

await getTeamServices({});

Workflows

Main Portfolio Assistant Workflow

The primary workflow manages the conversation flow:

1. Process user input with the LLM
2. Execute requested tools
3. Generate appropriate responses

const portfolioAssistant = new StateGraph(MessagesAnnotation)
  .addNode("llmNode", llmNode)
  .addNode("toolExecutionNode", toolExecutionNode)
  .addEdge("__start__", "llmNode")
  .addConditionalEdges(...)
  .compile();

Proposal Workflow

A specialized workflow for generating high-quality proposals:

1. Generate initial proposal
2. Review and provide feedback
3. Revise if necessary or approve

const proposalWorkflow = new StateGraph(PortfolioStateAnnotation)
  .addNode("generateProposal", ...)
  .addNode("reviewProposal", ...)
  .addConditionalEdges(...)
  .compile();

Security

The assistant implements several security measures:

1. Restricted Email: Can only send to a single hardcoded address
2. Topic Constraints: Only discusses Robin's team and services
3. Duplicate Prevention: Avoids sending multiple emails for the same request
4. Input Validation: Uses Zod schemas to validate all inputs

Helper Functions

Several helper functions support proposal generation:

• generateApproach: Creates a tailored approach based on project type
• allocateTeam: Assigns appropriate team members
• generateTimeline: Builds a realistic project timeline
• generateBudget: Allocates budget percentages

Usage Examples

Basic Query

const response = await runPortfolioAssistant(
  "Tell me about your mobile app development services"
);
console.log(response);

Project Proposal

const response = await runPortfolioAssistant(
  "I need a proposal for an e-commerce website with a budget of $15,000"
);
console.log(response);

Team Information

const response = await runPortfolioAssistant(
  "Who is on your development team?"
);
console.log(response);

Contributing

1. Fork the repository
2. Create a feature branch: git checkout -b feature/new-feature
3. Commit your changes: git commit -m 'Add new feature'
4. Push to the branch: git push origin feature/new-feature
5. Submit a pull request

License

This project is licensed under the MIT License - see the LICENSE file for details.

Contact

Robin Rathore - robinsingh248142@gmail.com Portfolio: https://robinrathore.dev LinkedIn: https://linkedin.com/in/robinrathore