Manager

Role & responsibilities

Preferred candidate profile

Perks and benefits

Role Purpose

The Agent Curator is responsible for governing, training, validating and continuously improving the portfolio of AI Agents deployed across the Purchase-to-Pay service.

The role acts as the operational bridge between:

• bp business stakeholders
• Wipro operations teams
• AI engineering teams
• Process design teams
• Knowledge Graph teams

The Agent Curator ensures agents:

• perform accurately
• operate safely
• remain compliant
• continuously improve
• deliver measurable business value

Unlike an AI developer, the Agent Curator does not build the underlying models.

Instead, the role owns:

"How agents behave in production and whether they are actually delivering the intended operational outcomes."

Key Responsibilities

1. Agent Performance Ownership

Own operational performance for all P2P agents including:

• Requisition validation agents
• Contract compliance agents
• Supplier support agents
• Invoice processing agents
• Payment control agents
• Exception management agents
• Reporting and analytics agents

Ensure each agent:

• operates within approved boundaries
• delivers expected business outcomes
• remains compliant with policies and controls

2. Agent Training & Continuous Learning

Identify opportunities to improve agent performance through:

• prompt refinement
• workflow optimisation
• exception analysis
• pattern identification
• human feedback loops
• knowledge expansion

Work with operational teams to capture:

• recurring exceptions
• missed opportunities
• escalation patterns
• quality concerns

and feed these improvements back into the agent design process.

3. Human-in-the-Loop Governance

Design and maintain clear rules for:

• agent decisions
• escalation thresholds
• approval workflows
• exception handling
• human intervention points

Ensure sensitive activities remain appropriately governed.

Examples include:

• payment-related decisions
• supplier master changes
• sanctions-related activities
• exception approvals
• policy breaches

4. Agent Validation & Certification

Own operational acceptance and sign-off of agents before production deployment.

Validate:

Accuracy

• Correct recommendations
• Correct actions
• Correct outputs

Risk

• Compliance adherence
• Control compliance
• Security considerations

Business Impact

• Productivity impact
• Service improvement
• User adoption
• Outcome achievement

Maintain a formal agent certification process.

5. Knowledge Graph & Data Readiness

Work closely with the Knowledge Graph team to ensure agents have access to:

• Contracts
• Price books
• Catalogues
• Supplier information
• POs
• PRs
• SESs
• GRs
• Policies
• Procedures

Identify gaps that impact agent performance and prioritise remediation.

6. Exception Management

Own the operational review of:

• Agent mistakes
• Hallucinations
• Incorrect recommendations
• Failed automations
• Escalation failures
• Compliance exceptions

Conduct root-cause analysis and coordinate corrective actions.

The focus is on:

preventing recurrence rather than simply correcting errors.

7. Business Stakeholder Engagement

Serve as the primary operational interface between:

• bp business users
• Procurement teams
• Operations teams
• AI engineering teams

Gather business feedback and translate it into:

• agent improvements
• new use cases
• updated workflows
• enhanced controls

8. Agent Adoption & Change Management

Monitor:

• agent utilisation
• adoption rates
• user feedback
• satisfaction
• business trust

Develop interventions to improve:

• user confidence
• adoption
• effectiveness

Key Deliverables

The Agent Curator will own:

• P2P Agent Inventory
• Agent Certification Register
• Agent Performance Dashboard
• Agent Improvement Backlog
• Human-in-the-Loop Governance Framework
• Agent Training & Validation Plans
• Agent Risk Register
• Agent Exception Logs
• Agent Adoption Scorecards
• Quarterly Agent Value Assessments

Critical Success Measures

Success will be measured through:

Agent Performance

• Agent accuracy
• Recommendation quality
• Successful task completion

Adoption

• User adoption
• Active utilisation
• Stakeholder satisfaction

Compliance

• Policy adherence
• Control effectiveness
• Audit readiness

Business Value

• Productivity gains
• Exception reduction
• Resolution speed improvements
• Outcome achievement

Continuous Improvement

• Reduction in agent-related incidents
• Improvement in agent accuracy
• Increased automation coverage

Required Experience

Essential

• 510 years experience in:
• Procure-to-Pay operations
• Shared Services
• Procurement operations
• Process improvement
• Automation
• Strong understanding of:
• Requisitioning
• Purchase Orders
• Invoice Processing
• Supplier Operations
• Payments
• Experience working with:
• AI copilots
• Workflow automation
• Digital transformation initiatives

Preferred

• Agentic AI operational experience
• SAP / Ariba / VIM experience
• Process Mining experience
• Knowledge Management experience
• Prompt Engineering exposure
• Generative AI governance knowledge

Behavioural Profile

The successful candidate will be:

• Curious about AI
• Process-focused
• Operationally minded
• Analytical
• Detail-oriented
• Strong communicator
• Comfortable challenging both technology teams and operations teams

Most importantly:

Able to translate business requirements into agent behaviours, and agent capabilities into business outcomes.