Three decades ago, the contact centre industry was bound to physical facilities. In these bustling, centralised hubs, rows of agents sat side-by-side, handling simple, straightforward voice calls within rigid brick-and-mortar constraints.
Workforce management (WFM) was a rudimentary administrative task, largely limited to basic time-and-attendance tracking and simple shift allocations. The operational landscape was uncomplicated, as Audrey William, Industry Analyst and founder of Crayon IQ, notes, “Twenty-eight years ago, contact centres were strictly physical facilities. It was simple and basic—agents worked in a single building, took calls, and digital channels didn’t exist. Back then, workforce management was nothing more than basic time and attendance.”
“Over time, the contact centre industry scaled from 200 agents to thousands, introducing outsourcing and complex multi-time-shift environments, workforce management had to evolve from simple allocation into a highly sophisticated operational mechanism.”
Currently, as artificial intelligence successfully automates routine, low-level enquiries, human professionals are left to handle only the most volatile, emotionally charged escalations. This shift has completely reinvented workforce management, moving organisations away from traditional historical data models toward sophisticated, multi-tiered short-interval forecasting.
This technological shift has given rise to entirely new roles such as AI supervisors, bot configuration specialists, and automated compliance analysts. It’s no longer just about tracking human hours, but about harmonising the collective strengths of humans and machines.
Modern WFM systems must now predict not just overall call volumes, but the exact duration, concurrency spikes, and seconds-long latency required for a seamless AI-to-human handoff. Mark Buckley, VP of Australian and NZ, Genesys, comments, “As customer experience have become more digital and omnichannel, workforce management has shifted from static planning toward real-time orchestration of resources across customer journeys. AI and automation are helping organisations to anticipate demand, improve workforce agility and make more informed decisions that enhance both service outcomes and employee engagement”.
The ecosystem of active AI agents operating within enterprise communication suites has experienced a massive 15x year-over-year increase, emphasising why modern WFM solutions must now forecast for both human shifts and automated bot capacity. Sam La Macchia, strategic advisor for WFM startup Real Numbers, comments, “The rise of cloud computing and machine learning techniques has enabled WFM platforms to evolve from slow batch processing to fast, parallel, optimal model execution. This is exactly what is required for real-time contact centre analytics and optimisation.”
COVID-19 and hybrid work
The most disruptive period for contact centres occurred during the COVID-19 pandemic, when the sudden transition to remote work forced a complete overhaul of traditional models. “COVID-19 was the toughest period the contact centre industry has ever faced. Global hubs like the Philippines and India—which serve critical markets across Australia, the US, and Europe—were hit incredibly hard as they scrambled to set up home internet facilities and infrastructure for workers overnight”, says William.
The crisis accelerated the emergence of modern WFM solutions that natively integrate:
- Flexible and hybrid scheduling: Built-in features designed for split locations and varying time zones. “Modern workforce management isn’t just about ensuring coverage anymore – it’s about managing flexibility. The best platforms on the market today have hybrid work, gig-style shifts, and agent autonomy baked directly into their DNA”, says William.
- Gig-style shifts and shift bidding: Mobile application features that allow workers to autonomously swap shifts or bid on hours based on personal availability and family commitments. “Features like shift bidding and automated swapping are no longer ‘nice-to-haves’—they are retention tools. Giving an agent an app where they can seamlessly bid on, swap, or adjust shifts around their personal lives shifts the power dynamic, turning scheduling from a managerial headache into a self-service benefit.”
These innovations balanced the workforce’s demand for flexibility with the operational necessity of strict compliance and real-time adherence.
The shift toward omni-channel communication—encompassing digital text, social media, and voice—has dramatically heightened the importance of skills-based routing. This transition is highly evident across Asian markets, particularly in Japan and South Korea, where social media channels are deeply integrated into daily communication.
In these regions, a customer journey frequently begins as a text interaction on platforms like WhatsApp before transitioning into a live voice call. Modern WFM scheduling systems must dynamically account for this context, ensuring that the human agents available at the exact second of transition possess the highly specific skill sets required—whether for specialised billing inquiries or high-friction product complaints that automated channels failed to resolve.
Managing the mixed human and AI workforce
As artificial intelligence increasingly automates lower-level, repetitive inquiries, human agents are left to handle highly complex, emotionally charged escalations. This shift makes forecasting and capacity planning significantly more intricate, requiring future WFM systems to manage a combined army of AI agents, chat bots, and human personnel. Buckley notes, “AI is transforming workforce management into workforce orchestration by helping organisations coordinate work across both human employees and AI-powered automation. Rather than simply automating tasks, AI helps organisations better anticipate demand, understand customer needs and optimise how work is distributed across the enterprise”.
William adds, “”The workforce management of the future isn’t just about scheduling people; it’s about forecasting an entirely new, blended ecosystem of voice AI, digital AI agents, and human professionals. This introduces a radical shift in planning capacity, routing logic, and algorithmic complexity that contact centres have never had to grapple with before.”
The emergence of dedicated AI operations Roles
The proliferation of virtual agents introduces an entirely new layer of operational overhead. “We are about to see the birth of entirely new job classifications within the contact centre that have nothing to do with taking traditional phone calls or answering WhatsApp messages. We are going to need AI Supervisors, QA Transcript Auditors, and Real-Time Interveners whose sole job is to monitor live bot conversations and step in the second a bot hallucinates or goes off the rails”, says William.
Organisations are introducing dedicated tasks and roles into their workforce plans, including:
- AI Supervisors & monitor analysts: Personnel tasked with watching live bot interactions to intercept hallucinations or off-rail conversations.
- AI configuration specialists: Staff dedicated to fine-tuning virtual agents.
- QA Reviewers for AI Transcripts: Quality assurance professionals dedicated specifically to auditing bot-driven historical data.
“Who schedules the people who manage the bots? These new roles—AI configuration specialists, fine-tuners, and supervisors—must be baked directly into the workforce plan as dedicated, highly specialised tasks. Contact centers are completely unaccustomed to scheduling for this layer of operational oversight.”
Buckley adds, “Predictive AI can anticipate demand shifts through forecasting and capacity planning, while workforce orchestration can dynamically adjust staffing priorities and routing decisions as conditions change. By leveraging workforce, interaction, routing and operational data from across the business, organisations can respond more effectively to fluctuating volumes, improve first-contact resolution and reduce customer effort”.
Forecasting and predictive models
To prevent breaking the customer experience during an AI-to-human escalation, contact centres are moving away from traditional historical data models. Advanced vendors deploy multi-scenario algorithmic bands to identify exact patterns where AI-to-human handoffs fail, allowing the system to continuously self-correct and buffer human capacity for unpredictable spikes.
La Macchia comments, “Traditionally, workforce management was underpinned by Erlang C queuing mathematics—a simple ‘forecast, schedule, and execute’ loop built on a fixed pool of agents. Today, those legacy metrics have been thrown a spanner in the works.”
Erlang C is a standard mathematical formula used in call centres and workforce management (WFM) to estimate the number of staff required to handle a specific volume of incoming calls within a desired service level. As contact centre operations have become increasingly complex, the effectiveness of this formula has become increasingly limited
“We have evolved from static scheduling into a new era of continuous learning, continuous sensing, decision support, and adaptive systems. The old way of hoping to smooth out seasonality or left-field events is no longer enough to meet desired service levels”, says La Macchia.
“The real friction point in the modern contact centre is the escalation and handoff”, says William, “When a bot realises a query is beyond its capacity and says, ‘Let me get you to a home loan expert right now,’ the WFM system must have already anticipated and reserved human capacity for that exact, highly specific spike. This isn’t a normal queue; it’s a high-stakes, real-time handoff that requires an entirely new approach to scheduling.”
“When an AI handoff happens, you aren’t just routing a call; you are routing context. The scheduling and forecasting models must be dynamic enough to account for these unpredictable human-in-the-loop escalations, ensuring the right specialist is sitting there waiting the moment the machine hits its limit.”
Synchronous voice AI vs. asynchronous chat
Forecasting text and chat interactions allows for a degree of latency; automated systems can handle multiple concurrent sessions, queuing inquiries until an expert is available.
Conversely, voice AI operates synchronously in real time. Latency must remain extremely low, and forecasting models must accurately predict not just standard inbound call volume, but the precise duration, timing, and concurrency peaks of automated handoffs to humans.
William advises, “”When we talk about leveraging historical data across channels, we have to recognise a fundamental architectural divide: voice forecasting is strictly synchronous, requiring real-time, instantaneous matching of supply and demand. You cannot forecast a live voice AI agent the same way you forecast asynchronous digital channels like chat or email, where time-buffers exist. The temporal math is entirely different.”
Modern AI workforce orchestration can drastically cuts down on manual tinkering. By continuously tracking service levels, staff numbers, and demand through non-stop forecasting, it stays ahead of the game. Buckley comments, “These systems can automatically reforecast demand, rebalance workloads, reallocate resources and help supervisors prioritise actions based on real-time conditions. By combining workforce intelligence and operational intelligence, organisations gain greater agility and resilience while maintaining service levels and employee productivity”.
He adds, “Rather than relying on manual schedule adjustments, workforce leaders can focus on strategic decision-making while AI helps optimise day-to-day operations. The result is a more adaptive workforce strategy that supports customer experience, employee engagement, employee flexibility and operational resilience, even during periods of rapid change”.
Automating compliance and Quality Assurance (QA)
Legacy quality assurance practices historically limited organisations to auditing a small fraction—typically 1% to 5%—of customer interactions via manual review. William comments, “”For decades, contact centres accepted a massive blind spot, manually reviewing a meager 1% to 5% of interactions and hoping they caught the systemic failures. Modern AI completely obliterates that limitation, giving us the power to pre-score and evaluate 100% of interactions instantly across every single channel.”
“Moving from a 1% random sample to 100% automated coverage fundamentally transforms QA from a reactive check-the-box exercise into a comprehensive operational radar. We are no longer guessing agent performance or compliance; we are measuring the entire operation in absolute terms”, she adds.
For highly regulated sectors such as banking, insurance, investment, and healthcare, real-time analytics act as an instantaneous risk-mitigation tool. AI systems monitor live feeds to immediately flag prohibited sales behaviours, vulnerabilities, or compliance omissions, such as a failure to state mandatory disclosures or secure proper customer consent.
While AI dramatically increases detection coverage and speed by prompting supervisors of active risks, final governance policy interpretation and regulator-facing decisions remain strictly human-led functions.
Competitive landscape of the WFM/WEM Market
The WFM/WEM market is undergoing a structural consolidation, according to William, shifting toward unified, single-stack platforms. She highlights the following
Legacy providers
- NICE: Continues to pioneer the space, recently launching its Workforce Empowerment Suite at the Nice World Conference in Orlando. The suite specifically addresses the governance, scheduling, and management of hybrid workforces comprising human agents and conversational AI.
- Verint & Calabrio: Traditional standalone heavyweights that have spent decades dominating WFM analytics and scheduling infrastructure.
CCaaS and cloud suite integration
Contact Center as a Service (CCaaS) providers are aggressively building or acquiring native WEM capabilities to eliminate the friction of third-party integrations.
- Genesys & Talkdesk: Actively embedding robust, native WFM and QA modules directly into their core communication suites.
- Five9: Expanding infrastructure to provide comprehensive management within a unified platform.
CRM and IT Service Management disruption
- Salesforce & Zendesk: Capitalising on the rise of autonomous AI agents by positioning themselves as single-pane-of-glass solutions. For instance, Salesforce’s Agent Force Contact Center strategy seeks to control the entire interaction, WFM, and QA layer within a single platform stack.
- Cisco: Recently introduced Webex Workforce Management alongside specialised QA tools, signaling a broader push to keep enterprise clients within the Webex ecosystem.
With major product releases rolling out globally, the broader market buyer’s guides detailing regional and APAC deployment perspectives are anticipated to publish before the end of June.