Quote generation tool on top of an ERP.
Bespoke product built on a client's ERP in Spain: cut quote preparation from 30-40 to 2-5 minutes through automation and AI.
A client in a technical business needed to improve quote generation times in their ERP (initially between 30 and 40 minutes per quote). They also wanted the simpler quotes to be handled by administrative staff, so that specialists only had to review them.
The challenge
Add a layer of value in terms of usability, performance and autonomy, without decoupling the product from their existing ERP.
The needs
- Optimise quote preparation times.
- Simplify the process so that administrative staff can take ownership of drafting, delegating only the review to specialists.
- Autonomy to add visual improvements to client deliverables (images with descriptions, typography, document structure).
- Generated quotes that are easily customisable.
Opportunities identified
- Leveraging the client's existing resources and knowledge to — with no cost increase and using familiar tools — improve the speed and quality of quotes.
- Gradual introduction of AI-based automation tools to minimise human error and reduce preparation/review times.
How it was approached
Rather than replacing the ERP or building a new product, the solution was built on top of the infrastructure and tools the client already used, minimising the transition cost and the learning curve. On that foundation, an automation layer was added — with AI support for the repetitive parts — that prepares the bulk of the quote and leaves specialists with only the final review. The result is a product that administrative staff can operate autonomously, with easily customisable templates and more polished deliverables, without decoupling from the original ERP.
Results
- Preparation times reduced from 30-40 minutes to 2-5 minutes.
- A negligible increase in fixed costs (around €10 / month, as there was previously no IT infrastructure in place).
- Improvements to documentation processes: everything goes where it should and is recorded as it should be, reducing human error.
Delivery time: 2 weeks.