20 pages
Glossary.
Short, plain definitions of the terms that come up when AI meets operations. Each entry gives the answer first, then how the idea works in a real process.
- AI agent workflowAn AI agent workflow is the defined path an AI agent works inside: its steps, the tools it may call, the points where it hands a case to a person or another system, the conditions that make it stop, and what it logs.
- AI deploymentAI deployment is the moment an AI system leaves testing and starts doing real work inside a live process: reading production data, writing to the systems of record and handing cases to people.
- AI guardrailsAI guardrails are the controls that keep an AI system inside what the business allows.
- AI proof of conceptAn AI proof of concept (PoC) is a short, bounded test of whether an AI system can do one specific job well enough to justify building it properly.
- AI-powered RPAAI-powered RPA is robotic process automation with AI models added at the points where a plain bot fails: reading documents, classifying requests, understanding free text or choosing between options.
- Back office operationsBack office operations are the internal functions that process, record and settle a company's work without facing the customer: billing, order processing, reconciliation, claims, records, procurement and compliance.
- Billing disputeA billing dispute is a customer's formal challenge to a charge on a bill or invoice, on the grounds that it is wrong, unauthorized or not owed.
- Business process reengineeringBusiness process reengineering (BPR) is the radical redesign of a process around the outcome it exists to produce, starting from a blank page instead of the current steps.
- Call deflectionCall deflection is the handling of a customer contact that would have reached a phone agent in another channel instead, usually self-service, chat or a proactive message.
- CRM data hygieneCRM data hygiene is the ongoing practice of keeping customer records accurate, complete and current as part of daily work.
- CRM data qualityCRM data quality is the degree to which customer records are fit for the work that depends on them, such as billing, service, renewals, routing and reporting.
- Intelligent automation (intelligent process automation)Intelligent automation, also called intelligent process automation (IPA), combines rule-based automation such as RPA and workflow software with AI that reads, classifies and decides, so a process can run end to end on messy inputs.
- Intelligent workflow automationIntelligent workflow automation is a workflow in which software makes some of the decisions a person used to make, such as classifying a request, choosing the next step or drafting a reply, inside limits the business sets.
- Process intelligenceProcess intelligence is the practice of building a continuous, data-based picture of how a process actually runs, using event logs from the systems it touches plus business context such as volumes, costs and outcomes.
- Process miningProcess mining is a family of techniques that reconstruct how a process actually runs from the event logs of its systems, where each record carries a case ID, an activity and a timestamp.
- Process redesignProcess redesign is changing how a process works, its steps, handoffs, decision points and roles, to get a better outcome.