Reads purchase orders
Reads part numbers, quantities, due dates and customers from fax, email and PDF orders.
NS-Order.
Automatically process orders arriving through different channels
AI reads purchase orders sent by fax, email or PDF, as well as each customer’s EDI screens, and creates order data with part numbers, quantities and due dates. Your staff simply review and approve it. The order acknowledgment is sent to the customer automatically, and the order moves on to the next step.
- Fax, email, PDF and EDI screens
- Phone orders supported (ask us)
FEATURES
NS-Order features at a glance
Reads EDI screens too
Reads each customer's EDI screen and puts those orders in the same order list as the rest—no more checking every EDI and retyping.
Phone orders too
Works with an ordinary phone line or with an AI phone service (ask us for details).
Registered after a check
What the AI reads lines up on the approval screen. Once staff check, fix and approve it, it goes into the order list.
Results in numbers
The results report shows orders processed, time saved versus manual work and how often staff corrected the AI.
CHANNELS
Orders from different channels, brought together
Whether an order arrives by fax, email, PDF, EDI or phone, AI reads it and it joins the same order flow. Once your staff check and approve it, the acknowledgment is sent to the customer automatically and the order moves on to the next step.
- FaxReads the image
- EmailExtracts the order from the body
- PDFReads the attached order
- EDIReads the order on screen
- PhoneTranscribes the call
One order flow
- 1Organized into part no., qty and due date
- 2Staff check and approve
- 3Acknowledgment sent automatically
- 4On to the next step
EDI support covers Web-EDI; for other EDI, please ask us. For phone orders, we can transcribe recorded calls or use an AI phone service (ask us for details).
CONFIDENCE
Reading accuracy shown as a confidence score
Along with what it reads, the AI outputs how confident it is. Low-confidence content is flagged with a warning and sent for a human check, and unreadable fields are left blank rather than guessed.
Order email from a customer
Confidence score95%
→ Straight to approval
Fax
Faded fax order
Confidence score62%
→ To approval with a warning
Phone
Hard-to-hear phone order
Confidence score41%
→ Blank fields, sent for review
NO MISSED ORDERS
Nothing slips through
A missed or duplicated order turns straight into late deliveries or overproduction. Every order that arrives is recorded so it can be traced later.
No duplicate intake
Each email is logged by its unique ID, so the same email is never processed twice. Emails you send yourself are ignored.
Originals are always kept
Fax images and received emails are saved as-is, even when reading fails, so the original order can always be checked.
Failures are recorded too
Orders that couldn't be processed stay on record as "failed", so you can see where they stopped. Nothing silently disappears.
Safety by Design
Three safeguards that bring peace of mind to the floor
“What if something goes wrong?” “I'm uneasy when I can't see inside.” “What happens if it stops?” The NS Series answers these real concerns from the floor with built-in safeguards—thoroughly removing the anxiety around AI. This is the foundation of the NS Series' peace-of-mind design.
Full process visibility
Check the AI's processing history in real time. No black box.
Switch to manual anytime
Even if the system fails, revert instantly to your traditional manual process—minimizing the risk of downtime.
Emergency stop switch
In an anomaly or emergency, halt all automation with a single tap—preventing unexpected trouble from spreading.
PRICE
Pricing
Single-function module
NS-Order
- Setup fee
- ¥500,000
- Maintenance (monthly)
- ¥10,000
* Prices exclude tax.
What maintenance covers
- On-site response to urgent failures
- Support by email and phone (up to 5 times a month)
- Regular system updates
LINEUP
Product lineup
Start with a conversation
Start with a free workflow assessment
We'll propose the best rollout plan for your shop-floor workflow. Why not start simply by hearing what it could look like?