How Nature-Inspired Engineering Shaped an Intelligent Workforce Management Solution

The primary objective of workforce management (WFM) is to optimize productivity and minimize risks. When used as a cutting-edge, intelligent software solution, WFM can contribute to the improvement of employee relations and customer experience (CX), ultimately leading to a favorable financial outcome for the business.

Scheduling tasks to manpower and expecting them to do their jobs optimally, like the traveling salesman problem, has long been questioned, researched, and analyzed to test for operational posterity. Not every person responds to a job assignment productively despite being qualified to do the job.

Typical roadblocks in workforce management

Ignoring employee psychometric scores, we observe that not every individual performs well in situations of high pressure and demand. Not everyone is welcoming to taking orders and commands. When a community or an eco-system is a self-regulating and self-conditioning entity that works together and cohesively for the common good, other community members can pitch in to cover up aberrations in an individual’s performance.

Biases of assignments, a lack of or poor pattern-matching of an agent to a domain/language/ switch, or predispositions of agents toward or against an account all add up to bad work productivity numbers. Add in the constraints of shifts, time slots, and time zones; these problems then keep getting multiplied and compounded. Furthermore, there are pressures of KPIs like reduced AHT numbers and increased CSAT and NPS scores. Agents, hence, work under immense duress to fruitfully comply with their schedules while being contributors.

Schedule adherence and tardiness via disciplinary and regulatory issues start to creep in, and payable but non-billable work activities take a hit. Non-payable work that promotes goodwill, like community outreach programs, goes for a toss, and agents only regard “on-call at-desk” work as a priority. Queue prioritization for handling calls also causes strain on the workers, thereby reducing the quality of their outputs, as reflected in their monthly audit scores.

• How can one solve this age-old problem of optimal schedules and optimizing workforce management through intelligent, quantifiable, and measurable resource allocation?
• How can businesses set standards, derive benchmarks, and formulate individual goals?
• How can a floor team leader assign duties optimally to meet an account’s demands while an RTA design is aimed at effective employee engagement rather than increased revenue generation?
• How can an organization have an informed rewards and recognition (R&R) program?
• How can compensation and benefits (C&B) plans evolve through employee assignments when “group-knows-better” thinking becomes the foundation for new team KRAs and “common good” becomes the new individual KPI norm?

Nature-inspired engineering in workforce management

To mitigate the above-mentioned problems, HGS has drawn heavily from nature-inspired algorithms, including the firefly algorithm, artificial bee colony algorithm, and ant hill colony algorithm, to render intelligence to the scheduling problem.

In addition, several modern machine learning methods have been introduced to the study, including archetypal analysis, anomaly detection, abnormality detection, novelty detection, market basket analysis, association rule mining, and RFM & RFD analysis to impute intelligence to the entire floor management activity of team leads and RTAs in the contact center paradigm.

Metrics have evolved and are now tangible and deterministically mensurate for a periodic, holistic, and unbiased review of account management, audit trails, agent performance scoring, and individual/group appraisals, thereby moving away from traditional scoring and review systems that are based on speculative goals, subjective opinions, and stochastic surveys.

How do nature-inspired workforce management techniques work?

As part of mimicking the behavior of beehive colonies, firefly mating procedures, or ant colony production lines, HGS imparts artificial intelligence (AI) to the models built on real-world data that assumes role-based assignments for community good rather than individual brilliance.

The motto, being a team of good, loyal, and dedicated players who listen to instructions and take commands, always wins over a team of stars with egotistical individuality and self-centric decision-making. However, individual brilliance is not ignored and is aptly rewarded via equitable weighting mechanisms, where applicable.

Additional knowledge from the aforementioned techniques for group segmentation and micro-targeting helps understand the “individual” aspects of the crowd to cater for more direct attention. It is easier to treat a person in a crowd for unacceptable behavior based on group standards, much like how the animal kingdom corrects an erring member in pride.

Industry applications of nature-inspired WFM

The scope here is not just limited in or to a contact center scenario. Even in mundane activities like assigning household chores, shift assignments for doctors/nurses in hospitals, matching teachers to grades, or patrolling policemen, night guards, or delivery personnel, scheduling seamlessly finds connections in the modern age across the spectrum wherever a supply chain or goods/service transfer transaction is involved.

Functioning like a chain of ants, a flock of bees, or a swarm of fireflies provides for group intelligence and near real-time feed-forward or backward propagating errors and warnings (read pheromones-based biochemical messages in the animal world), something very akin to deep learning techniques.

The HGS approach to workforce management

Prior market research identified substantial gaps in data science-infused intelligence in WFM solutions, both custom-made as well as in COTS (commercial-of-the-shelf), which HGS addressed with its distinctive and unparalleled home-grown solution that boasts –

• Innovation
• In-house solutions with no external dependencies
• Cloud-enabled and optimized
  • Scalable and repeatable
  • Modular, trainable, and customizable
  • Plug-and-play
• Offline or real-time
• Secure
• On-premise or cloud deployment

HGS’s data science lab introduced these concepts to the WFM team spread across seven geographies, catering to a multitude of clients across numerous domains. It propagated the scope of AI through these approaches via a prototype design based on a BFSI account, currently served by the Jamaica team in a voice-only, English-only 24x7x365 support system.