The problem
Enquiries arrived by email as free text. Customers described products in everyday language that did not correspond exactly to the technical names and codes in the catalogue.
For every enquiry, the sales team searched the catalogue, found the correct SKU and prepared the quotation document by hand. Across tens of thousands of SKUs, the search often took longer than the underlying sales task.
Basic text search was not enough: multiple names for the same product, similarly named but distinct products and incomplete enquiries led to incorrect matches.
What we set out to improve
- Time from receipt of an enquiry email to completion of a quotation draft
- Rate at which products described in free text are matched to the correct SKU
- Time the sales team spends searching the catalogue by hand
How the system works
- 01 Receive the enquiry email and its attachments
- 02 Parse the text and extract the requested items
- 03 Convert everyday descriptions into product concepts
- 04 Find candidate SKUs in the catalogue
- 05 Calculate a confidence score for each candidate
- 06 Send uncertain matches below the threshold to a person for review
- 07 Prepare the quotation document
- 08 Record the sales team's final approval
Where people stay involved
The system stops after preparing the quotation document. The sales team reviews product matches, corrects any errors and decides whether the quotation may be sent.
Results
The project team reported that quotation preparation seemed markedly shorter than in the previous manual process.
This was not measured through a time study; it is an observation by the project team.
Data and security
Access to enquiry emails, attachments, catalogue data and quotation records is limited by role. Uncertain matches and every approval decision remain traceable to their source records.
Who this suits
A good fit
- B2B sellers with catalogues containing thousands of items and enquiries arriving as free text
- Manufacturers of technical products whose catalogue terminology differs from customer language
- Companies where quotation preparation occupies a substantial share of the sales team's time
Not a good fit
- Businesses with a small, stable catalogue, where a conventional form and price list are more suitable
- Businesses whose enquiries already arrive in a structured format with product codes, so no matching is required
- Businesses with outdated or inconsistent product data, which need to organise their data before applying AI
Frequently asked questions
Is ERP integration required?
No. The workflow can prepare a quotation draft from the available catalogue and enquiry data, although integration may reduce manual data transfer where a suitable interface exists.
How are incorrect product matches prevented?
The system ranks candidate SKUs, applies confidence thresholds and sends uncertain matches to the sales team. A quotation is not sent until a person has reviewed it.
Can human approval be removed?
It should remain in place. Product descriptions can be incomplete or ambiguous, so the sales team must make the final decision before a quotation is sent.
How well does the system understand free text?
Performance depends on the quality of the catalogue, the terminology used by customers and the available examples. Ambiguous descriptions are presented for human review rather than treated as certain matches.
What determines the implementation time?
It depends on catalogue size and consistency, the formats of incoming enquiries, the quotation template and the integration points that are available.
Project led by:DijitalPi
✦ PI ASSISTANT · ARTIFICIAL INTELLIGENCE
onlineAsk the DijitalPi AI Assistant your question
No need to fill in a form and wait for a reply — write your question and let Pi, trained on DijitalPi's 20 years of know-how, answer within seconds.
- Instant answers, 24/7
- Service-specific context — no generic replies
- Can connect you to a free consultation if you wish
Pi Assistant answers are for information purposes; for a proposal get in touch.
ASK AI ABOUT DIJITALPI
Let an AI explain what DijitalPi does.
Opens your chosen assistant with a ready research prompt. It reads the site live and answers.
Cookie preferences We use cookies to improve your experience and to analyse site traffic anonymously. Details: Cookie Policy · Privacy