AccuBid to Word Automation System - PLEASE DON'T USE GENAI FOR BIDDING. I WILL PRIORITISE ACTUAL BIDDERS

Job ID: 39555438

Budget: ₹1,500 – ₹12,500 INR

## Project Description: AccuBid-to-Word Conversion System

### Project Goal
Develop an AI-powered system that converts variable-complexity AccuBid outputs into standardized Word documents using **OpenAI Assistants API exclusively**. The solution must include tool calling for document generation, email automation, calendar integration, and Excel updates.

### Core Requirements
1. **OpenAI-Centric Implementation**
- Use **Assistants API** for all GenAI functions (no LangChain)
- Leverage built-in vector store for RAG (limit: 10K files)
- Implement native function calling for tools

2. **Key Functionality**
- **Dynamic Document Generation**: Handle projects from 50 to 5,000+ line items
- **Tool Calling Suite**:
- Generate Word docs via `python-docx`
- Email documents using SMTP
- Create calendar events via Microsoft Graph API
- Update Excel sheets through `openpyxl`
- **Auto-Updating RAG**: Hourly knowledge refreshes via Supabase Edge Functions

3. **Technical Stack**
```python
# Mandatory Components
Frontend: Streamlit
Backend: Supabase (PostgreSQL + Edge Functions)
AI Engine: OpenAI Assistants API (exclusively)
Tools: python-docx, openpyxl, smtplib, msal
```

### Specific Deliverables
| Task | Details |
|------|---------|
| **Streamlit UI with Supabase auth** | User-friendly frontend for data upload and tracking |
| **AccuBid data parser (Excel/API)** | Ingest and preprocess project data |
| **OpenAI Assistant setup** | Configure with few-shot prompting and function calling |
| **Tool calling implementation** | Automate Word, email, calendar, and Excel workflows |
| **RAG vector store initialization** | Use OpenAI’s built-in store for retrieval-augmented generation |
| **Scheduled RAG updates** | Supabase cron job for hourly refreshes |
| **End-to-end testing** | Ensure all workflows are robust and reliable |
| **Deployment documentation** | Clear instructions for setup and use |

### Essential Features
- **Adaptive Formatting**: Handle vastly different project scales through chain-of-thought reasoning
- **Automated Workflow**:
```mermaid
graph LR
A[AccuBid Input] --> B(OpenAI Assistant)
B --> C{Tool Calls}
C --> D[Word Docs]
C --> E[Emails]
C --> F[Calendar]
C --> G[Excel]
```
Related categories: Python RESTful API Google Firebase OpenAI Streamlit