◎ CASE / 002◎ AI AGENT◎ JP◎ 2024● LIVE
JP Mid-size Manufacturer
JP Quote Automation
JP Quote Automation
Parses RFQ emails, queries price DB, auto-generates Excel quotes.
ROLE
Automation Workflow Build · Japan Localization
PERIOD
2024.09 — 2024.12 (3 months)
MARKET
JP
HEADLINE
-87% · Quote generation time
(01) Challenge · The problem
“Starting point”
A Japanese SMB manufacturer received 40–60 RFQ emails per day, all in Japanese. Sales staff read the specs, looked up prices in Excel, built PDF quotes by hand, and replied — averaging 4 hours per quote. Tight-deadline requests meant overnight work.
(02) Solution · How we solved it
“Approach”
GPT-4 auto-extracts specs (material, quantity, lead time, dimensions) from Japanese email bodies and attached drawings. After querying the PostgreSQL price database, it generates Excel and PDF quotes from templates. Sales staff click once to confirm and the quote sends automatically. Management approval emails are automated too.
(04) Outcomes · Results
Proven by numbers.
-87%
Quote turnaround
From 4 hours to 30 minutes on average (review included)
40–60 cases
Daily volume
Same load, no more overtime or weekend work
100%
Japanese accuracy
Trained on keigo and industry-specific vocabulary
2×
Quote coverage
Auto-handles requests previously dropped under deadline pressure
(05) Stack · Tools used
Trusted tools only.
01 · AI
- ▸GPT-4
- ▸Vision API (drawing OCR)
- ▸Python
02 · Backend
- ▸FastAPI
- ▸PostgreSQL
- ▸Redis
03 · Automation
- ▸n8n (mail trigger)
- ▸Excel.js
- ▸PDFKit
04 · Integrations
- ▸Gmail API
- ▸Chatwork
- ▸Internal ERP
▸ Related · AI AGENTALL PROJECTS →
START / FREE CONSULT · NDA OK