Spaces:
Runtime error
Runtime error
File size: 3,746 Bytes
46f2cb3 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 |
# Quick Start: Product Design RAG
## Setup (One-Time)
### Step 1: Copy Documents to Modal Volume
```bash
# Documents should already be uploaded, but if needed:
modal volume put mcp-hack-ins-products \
docs/tokyo_auto_insurance_product_design_filled.md \
docs/tokyo_auto_insurance_product_design.docx
```
### Step 2: Index the Documents
```bash
# Option A: Using the CLI tool (auto-detects venv)
python3 query_product_design.py --index
# Option B: Using helper script (activates venv automatically)
./run_with_venv.sh --index
# Option C: Activate venv first, then run
source venv/bin/activate
python query_product_design.py --index
# Option D: Direct Modal command (if venv is activated)
modal run src/modal-rag-product-design.py::index_product_design
```
This will:
- Load the markdown and Word documents
- Split them into chunks
- Generate embeddings
- Store in ChromaDB collection `product_design`
**Expected time**: 2-5 minutes
## Querying
### Option 1: Interactive Mode (Recommended)
```bash
python query_product_design.py --interactive
```
Then type your questions:
```
❓ Your question: What are the three product tiers?
❓ Your question: What is the Year 3 premium volume?
❓ Your question: exit
```
### Option 2: Single Query
```bash
python query_product_design.py --query "What are the three product tiers and their premium ranges?"
```
### Option 3: Direct Modal Command
```bash
modal run src/modal-rag-product-design.py::query_product_design \
--question "What coverage does the Standard tier include?"
```
## Example Questions
### Product Features
```bash
python query_product_design.py --query "What are the three product tiers and their premium ranges?"
python query_product_design.py --query "What coverage does the Standard tier include?"
python query_product_design.py --query "What are the unique features of TokyoDrive Insurance?"
python query_product_design.py --query "What add-on services are available?"
```
### Financial Projections
```bash
python query_product_design.py --query "What is the Year 3 premium volume projection?"
python query_product_design.py --query "What is the target loss ratio for Year 2?"
python query_product_design.py --query "What are the break-even projections?"
```
### Market & Strategy
```bash
python query_product_design.py --query "What is the target market size in Tokyo?"
python query_product_design.py --query "Who are the main competitors?"
python query_product_design.py --query "What are the key value propositions?"
```
### Technical Requirements
```bash
python query_product_design.py --query "What are the technology requirements?"
python query_product_design.py --query "What is the claims processing workflow?"
python query_product_design.py --query "What mobile app features are planned?"
```
### Compliance
```bash
python query_product_design.py --query "What are the FSA licensing requirements?"
python query_product_design.py --query "What is the minimum capital requirement?"
python query_product_design.py --query "What data privacy requirements apply?"
```
## Troubleshooting
### "No documents found"
- Make sure documents are in the Modal volume
- Check: `modal volume list mcp-hack-ins-products`
### "Collection not found"
- Run indexing first: `python query_product_design.py --index`
### Slow queries
- First query may be slow (cold start)
- Subsequent queries should be faster (warm container)
## Next Steps
1. **Test with your questions**: Try asking questions your team would actually need
2. **Create query templates**: Build common queries for different use cases
3. **Integrate with workflows**: Use RAG to extract requirements for development tickets
4. **Share with team**: Let stakeholders query the document themselves
|