January 20, 2026
KPMG and Uniphore Move AI Agents Out of Pilots and into Regulated Industries
KPMG has partnered with Uniphore to develop and deploy AI agents built on industry-specific small language models, with AI being applied in regulated enterprises beyond small-scale trials. The agents will support both internal KPMG teams and client-facing use cases.
The collaboration brings together KPMG’s domain knowledge and experience with small language models (SLMs) and Uniphore’s Business AI Cloud, a platform that supports agentic AI in complex business environments. The focus is on sectors where governance, security, and compliance are essential, including banking, insurance, energy, and healthcare.
Delivery at Scale
While many large organisations still use AI mainly in pilot projects, KPMG says this partnership aims to make it part of how work actually gets done.
Under the model KPMG is developing, consultants are trained to advise on strategy, design, deployment, and governance of AI agents. The aim is to combine human judgment with AI execution, allowing expertise that is usually locked in documents or individual teams to be reused and scaled.
Uniphore’s Business AI Cloud acts as one of the foundations for this approach. It allows KPMG to encode regulatory frameworks, institutional knowledge, and process playbooks into small language models that are specific to an industry or function. These models can then power AI agents across areas such as procurement, finance, workforce optimisation, claims management, and customer experience.
This work is supported by what KPMG describes as an “SLM factory” model. In practice, it means taking repeatable knowledge work and turning it into governed AI systems that can be reused across clients and sectors, rather than rebuilt each time.
Procurement as a Proving Ground for AI Agents
The collaboration focuses on procurement and contracting, where manual processes often slow reviews and increase risk.
The AI agents classify high-value contracts, compare terms with approved standards, extract obligations, flag risks, and route exceptions to human reviewers within existing procurement workflows, helping shorten review cycles, reduce revenue leakage, and improve risk oversight.
Importantly, these agents handle repetitive analysis and pattern matching, while complex or sensitive decisions remain with people.
“We are thrilled to align with Uniphore’s vision for AI as a transformative force for business as we focus on helping clients move from AI experimentation to real operational value,” said Prasad Jayaraman, Advisory Principal at KPMG.
Built for Enterprise Data Reality
Data remains a major obstacle for AI in large organisations, where information is fragmented, hard to access, and subject to strict regulatory controls. KPMG and Uniphore aim to make AI agents work directly with platforms such as Databricks and Snowflake, allowing them to operate on governed, production-grade data without data migrations or parallel systems.
The announcement aligns with the opening of the World Economic Forum Annual Meeting in Davos, where leaders from both organisations are meeting with clients and partners to discuss how business AI is entering its next phase.
