Pricefx adds bring-your-own AI to pricing platform

Pricefx has introduced a bring-your-own AI feature for its pricing platform, allowing customers to connect an approved enterprise AI assistant to the software.

The feature lets pricing, finance, and commercial teams ask questions in everyday language about live pricing data, transactions, and models through tools their employer already uses, such as Microsoft Copilot, Claude, or ChatGPT.

Pricefx said the approach is based on the Model Context Protocol, an open standard that links AI assistants with business software. Instead of requiring a proprietary integration, the platform is being opened so customers can use a wider range of assistants and switch if their internal AI strategy changes.

Teams can query pricing recommendations, customer margins, discounts and quote data without moving to a separate assistant dedicated to pricing work. Users can also examine the logic behind configured optimisation models and test how changing an input would affect an outcome.

In one example outlined by Pricefx, a pricing manager could ask an assistant to identify pricing patterns in a segment, explain what is driving a customer's margin and estimate the effect of a proposed discount. In another, a commercial lead could use an approved assistant before a renewal or quote to compare options and identify risks or trade-offs for review.

Open approach

The announcement reflects a wider trend among business software groups to work with AI tools already sanctioned by large companies, rather than asking customers to adopt another standalone assistant. For corporate users, that can reduce training demands and make it easier to fit AI features into existing governance and procurement frameworks.

Ronak Sheth, Chief Executive Officer at Pricefx, framed the release as an extension of that approach to software integration.

"Enterprises have already chosen the technology they trust, and we believe in working with those choices, not asking customers to start over. We've spent a decade meeting customers inside the systems they already run: SAP, Salesforce, whatever their landscape looks like. This is that same commitment applied to AI. Now they can connect the assistant their company has already approved and put it to work on real pricing decisions. Flexibility is how we've always earned trust, and that doesn't change just because the technology does," said Ronak Sheth, Chief Executive Officer at Pricefx.

He said a connected assistant can inspect how pricing works inside the system, including optimisation models, the logic behind recommendations and relationships between customers, products and quotes. That gives users more context than a general-purpose assistant working from an exported report or file, Pricefx argued.

Human oversight

Pricefx also emphasised that the AI assistant is not intended to make pricing decisions on its own. Under the set-up described by the company, the assistant can read information and prepare drafts, but actions remain subject to user permissions and the same approval processes companies already use.

That structure is intended to preserve governance in pricing, a sensitive function for manufacturers and distributors managing margins, customer contracts and discounting. By keeping a person in the loop, Pricefx is positioning AI as an analytical layer rather than a replacement for commercial judgement.

Sheth highlighted that balance between access and control in a second comment.

"Bringing in an assistant people already trust is the easy part. Keeping the pricing logic, the governance, and the accountability exactly where they belong is the part that makes it work in an enterprise," said Sheth.

Potential uses include understanding the basis of a system recommendation before acting, investigating product-level margin detail ahead of a customer renewal, testing the effect of discounts on margin, reviewing the structure of an optimisation model and monitoring overnight market shifts.

The feature will first be offered through an Early Adopter Program before a wider release. The staged rollout suggests Pricefx is using an initial customer group to test how the integration performs in live pricing environments, where data access, audit trails and approval controls are likely to be closely scrutinised.

Pricefx operates in the business-to-business pricing software market, where suppliers are increasingly adding AI-based interfaces to help users navigate complex data and recommendation engines. In that context, the decision to support existing enterprise assistants rather than launch a separate one marks a distinct product choice.

The feature is designed to let business users interrogate pricing data and model outputs more directly, including those who do not routinely work with pricing models or dashboards.