Capita launches AI Catalyst Stack in drive to be first AI-Enabled BPO

Embedding agentic AI into the fabric of business process outsourcing to deliver faster, smarter, and more trusted services across the public and private sectors.

Capita plc today announced the launch of The AI Catalyst Stack, operationalising AI at scale and underpinning its ambition to become the first AI-enabled business process outsourcer (BPO).

The AI Catalyst Stack leverages Capita’s operational expertise and hyperscaler partnerships to embed AI in mission-critical services, driving faster transformation, reducing costs, and improving outcomes for citizens and customers.

“AI is not an upgrade - it’s the cornerstone of our operational evolution,” said Adolfo Hernandez, CEO of Capita.

“With The AI Catalyst Stack, we’re embedding intelligence into the fabric of how we serve society - safely, repeatably, and at scale. This positions Capita as the first BPO where AI doesn’t just support work, it does the work, freeing people to focus on empathy, creativity, and governance.”

From traditional BPO to agentic BPO

For decades, BPO has been powered by people; The AI Catalyst Stack transforms it into a human plus agent partnership model. AI agents can now coordinate end-to-end workflows, while human experts oversee quality, compliance, and empathy-driven interactions.

This “agentic BPO” model, pioneered within Capita’s own operations as client zero, is already delivering AI-enabled value and measurable results in live operations:

Assessment services: AI agents guide and audit decision-making, cutting validation times and improving audit accuracy.
Public safety and fire services: Agents automate compliance checks and reporting, improving risk visibility and consistency.
Procurement and project delivery: Agents accelerate tender evaluation and standardise programme delivery.
Customer experience: AgentSuite integrates AWS Connect, Amazon Lex, and CapitaContact to drive 17–20% faster handling times, up to 35% productivity gains, and double-digit improvements in customer satisfaction.

Built for public trust, engineered for enterprise scale

As the UK government moves from “AI optional” to “AI expected”, The AI Catalyst Stack provides the safe, auditable foundation required by public-sector standards. Its embedded policy packs, data-handling controls, and human-in-the-loop oversight enable Capita to deliver AI outcomes that are transparent, ethical, and compliant by design.

The AI Catalyst Stack aligns with the UK Government's recent announcement on AI Growth Labs, which will enable faster innovation and reduce bureaucracy safely.

“Our mission is to become the first business process outsourcer where every service is AI-enabled by default,” said Sameer Vuyyuru, Chief AI and Product Officer at Capita.

“We’re building agentic AI directly into workflows — not as bolt-ons, but as the engine of delivery. That means clients see measurable value in a matter of weeks.”

Delivering better outcomes for clients and society

By embedding AI into how it delivers services, Capita is driving tangible business and societal impact:

Faster change: From bespoke builds to repeatable AI products, reducing delivery cycles by 60–70%.
Lower cost-to-serve: Evidence-led automation and shared data insights.
Smarter decisions: A semantic data layer turns raw information into trusted intelligence.
Empowered people: Employees are augmented, not replaced — equipped with tools that elevate judgment, empathy, and service quality.

The AI catalyst stack specifications

The AI Catalyst Stack is a modular architecture which leverages the innovative AI technology of our hyperscaler partners, and is designed to deploy and scale AI safely across regulated environments:

Process observability — Creates a living digital twin of business operations, surfacing inefficiencies, compliance risks, and automation opportunities.
AI build and deployment platforms — Combines low-code environments for rapid innovation with pro-code toolchains for complex workloads, underpinned by responsible-AI guardrails.
Orchestration and integration — A secure multi-cloud fabric connecting legacy applications, SaaS platforms, and AI agents through event-driven APIs (Application Programming Interface) and future proof MCP/A2A (Model Context Protocol/ Agent-to-Agent) protocols.
Data management and intelligence — A governed semantic layer enabling unified MI/BI (Management Information/ Business intelligence), real-time analytics, and machine-learning pipelines.