Jimmy S.

We work in the transactional quality management domain within a call center setup for the travel industry. Maintaining service excellence through structured audits, dissemination of product updates in a timely manner, refresher trainings targeted to key individuals, and performance insights is our focus We keep a close watch on performance trends, study the process health, and publish various reports, key insights on findings to leadership, helping them with data-driven decisions. Our day-to-day operations use tools like Excel, Smartsheets, PowerPoint, Sway, & Power BI, this helps us manage data efficiently and make real-time insights into process performance. Yes, traditionally a process can be improved by Lean/Six Sigma that focuses on Eliminating waste, reducing variation, and improving process flow These are powerful, but they assume the process itself is fundamentally sound and only needs tuning. However, AI introduces capabilities that can fundamentally change the nature of the work, not just optimize it. While these are traditional and powerful ways to fix and straighten processes, they assume that the process itself is fundamentally sound and only needs tuning. However, with the introduction of AI, various capability avenues open that can fundamentally change the day-to-day nature of work, not just limiting it to optimization alone. Automating day-to-day cognitive tasks Our team conducts audits, disseminates updates, and identifies training needs; these usually involve pattern recognition, decision-making, & communication. These tasks can be handled by an AI: Natural Language Processing (NLP) can analyze call transcripts for quality. Generative AI can draft personalized refresher content. Predictive analytics can forecast training needs before issues arise. Helping with real-time decision-making Currently used Power BI showcases and is focused on visibility, but AI can go further: Prescriptive analytics can suggest actions based on trends. Anomaly detection can flag outliers in real-time, not just report them. Scalability and Adaptability Upon deploying and fed with a larger database, AI systems can adapt to new client updates, changes in the business rules, & evolving customer expectations quicker than manual dissemination and training cycles. Suggestive vision on re-imagination for AI-powered transformation: Quality Audits conducted by an AI Deploying speech-to-text & natural language processing to automatically audit calls/transactions. Score transactions based on sentiment, compliance, and resolution quality. Flag calls/transitions for human auditor’s review only when it is necessary. Dynamic Knowledge Management Introduce AI chatbot or AI assistant trained on client updates & SOPs. This will be available 24/7 to answer agents'/processor's queries. The database will be automatically updated, and the same will be disseminated with the new product info to the pre-defined group. Predictive Training Needs The introduction of machine learning will help in analyzing audit scores, test results, and call/transaction metadata. Predict which processors are likely to need refresher training. Auto-schedule microlearning modules tailored to individual gaps. AI-Augmented Reporting These dashboards will not only show performance trends but also explain them. Natural language summaries of performance for senior executives/senior leadership Conversational Interfaces for Leadership Senior leaders interact with a voice or chat interface to ask: How did the team in Boston perform last week? What are the top 3 training gaps across delivery sites

In my opinion, in a hospital (that too a busy one) to check and track patient wait times there would be multiple sources to feed-in the data we are looking for. A patient's viewpoint, a hospital staff's view, and finally a third party's viewpoint. All these viewpoints may differ in experience and priority. for example: 1. For a patient there may be multiple wait time segments - waiting the lobby, waiting in the examination room, waiting in the treatment room etc. 2. The hospital staff like doctors, nurses, receptionists, facility management folks, etc. would have their own observation from their point of view as to which section or segment of someone attending to the patient is causing a hindrance/ obstacle 3. A neutral party like ours will have a different view altogether from what we see, when the data is collected with date and time at each section/ station right from the reception (lobby area), waiting in the examination room and finally attended by a doctor to further collecting medicines from the pharmacy. Once these various readings are taken a consolidation would give us a wider factor that contribute to the wait times. Tips, tricks, guidelines: * All numbers/ time slots need to be counted in one unit (e.g., minutes only, hours, hours only, seconds only) * Sampling needs to be done - time of the day, number of sample patients, days of the week etc. * A standard process simulation should be done with all operators/ data collectors so that no to less room for error even at the data collection phase * Double check/ review the reviewer how all operators are collecting the data - whether they are following a standard process as discussed in the above pointer * Conduct a mini simulation of the outcome/ observation and share with the hospital staff before considering the same with a larger data for the project to avoid any outliers/ missed out steps/ etc.

Q. Key Risk Indicators (KRIs) be used to manage a process? Yes, Key Risk Indicators (KRIs) can be used to manage a process. In cases we want to ensure resilience, compliance, & proactive risk mitigation. KRIs are not just for risk reporting—they can be utilized to work in tandem to operational workflows Q. Why do companies tend to prioritize Key Performance Indicators (KPIs) over KRIs for process monitoring? To detect early warning signs of process breakdowns. To trigger preventive actions before issues increase/ multiply, escalate. Support compliance with internal controls & external regulations Complement KPIs by suggesting a risk-adjusted view of performance. e.g., review the reviewer Spot Checks, Audit Readiness, PBI Reporting etc. KPIs are easier to quantify and communicate KPIs align with strategic and operational goals KRIs are leading indicators KRIs require contextual analysis and may not show immediate value unless a risk materializes Usually Organizational culture favors performance over prevention Q. Discuss the benefits and limitations of using KRIs with examples. 1. Benefit: Early Warning System KRIs help you spot trouble before it becomes a big problem. Example: In a factory a workman observes a rise in machine vibration levels; it can fix the issue before a breakdown. 2. Better Decision-Making: It can help managers/team leaders to make smarter choices by showing where risks are growing. In a BFSI setting, a bank back end notices more failed login attempts, leading to tightening security before a system breach.

Business Process Re-engineering can be utilized to address an outdated legacy system or classical way of performing transactions and carrying out business. For example, there was a period not long ago when airlines used legacy reservation systems that were largely people-dependent, with little to no automation. However, we’ve seen several mergers between airline companies where the best of both worlds was utilized to streamline, modernize, and automate their booking and ticketing processes. This even extended to the introduction of AI in recent times for handling cancellations and rebooking affected passenger itineraries. This was a complete overhaul from the traditional way of conducting day-to-day business. Lean Six Sigma has played its role at each of these tollgates by reducing variation and removing waste to improve these processes and their supplementary functions. While BPR brought about radical change in the business, LSS contributed by enhancing efficiency, productivity, quality, and customer satisfaction. In conclusion, although BPR and LSS may seem like rivals to a certain degree, we cannot deny the fact that they can be deployed and implemented in tandem.

I have some thoughts on how orienting, training, communicating, and involving team members would lead AI to earn trust: * To inculcate transparency, workshops need to be held where we demonstrate how AI works on customer (internal and external) queries and generates responses from available knowledge bases and from the web. * When deploying AI in a particular department, the inputs and feedback from team members need to be gathered . Create a channel or forum for them to share issues, improvements, and suggestions. This channel/forum should be looked at by Project Managers to review and act on the inputs from team members * Training sessions and plans to educate the team should be created and implemented. Once the team understands the working of AI and the value they bring, they will be more confident and shall trust the technology * Another way to earn and build trust for AI among the teams is by ensuring AI applications are built on the foundation of privacy and data security and that they comply with such policies and relevant regulations. * Communicating and showcasing tangible benefits produced by AI in day-to-day tasks. Making everyday tasks easier or more efficient would lead to earning trust in AI. e.g., improved efficiency, reduction in the time taken to resolve customer queries, increasing accuracy in data entry etc., will

There are couple of examples I would like to suggest: AI can be trusted - * Currently we have an AI deployed that addresses employee queries on HR policies at the first/ ground level. It provides guidelines and breaks down the complicated appearing policy terminologies in easy to understand language. AI can also re-direct/route random/ general queries regarding IT, HR, Admin an Facilities to the relevant team/s to accentuate timely and efficient query resolution * AI can also be deployed for inventory management and ordering (from a pre-defined vendor) of freebies/ goodies that we distribute to recognize outstanding employees for implementing ideas and fusion of Ideas, also for continuous improvement projects that are deployed and implemented. AI cannot be trusted - * On the other hand an AI intervention may not be trusted fully in the areas of discussing a feedback or the outcome of an audit that was conducted This is because the outcomes of the audits we conduct has severities (major, minor, observations and areas of improvement). There is a reasoning and dialogue that is required between the auditor and auditee to come to an agreement on the audit outcome and the way forward towards mitigation of the audit findings. At this moment an AI cannot be trusted to conduct this kind of complex task in an efficient manner as a human auditor would do.

Background: I am part of a team that facilitates Business Process and Excellence in the organization. One major wing of our team is to conduct audits for operating and support teams across the organization. These audits are based on the expectations of CMMI, Quality Management System and COPC (Customer Operations Performance Center). We have a team of auditors who conduct varied audits (Business Unit, Enabling and Top Management). Planning, scheduling and utilizing the auditors' skills, experience and availability to conduct these audits in an efficient manner is at times is a challenging task and requires lot of effort and time. Suggestion: AI can definitely be a valuable contributor across all phases of the audit lifecycle: Pre-audit Audit planning, scheduling, aligning the auditor). AI can study various aspects - annual target of audits to be conducted, plan and schedule the audits as per auditor skill matrix, experience and availability Sensitizing/ orienting the auditee team on the expectations (Data, team involvement and commitment, timelines etc.) - The very moment the audit schedule and auditor alignment is ready, the AI can send out automated email to the Auditee team and copy interested parties with an Audio/ Video version of the Audit Orientation presentation which is currently involves the auditor's time and non-auditing effort During the Audit Monitor audit progress and timelines cadence - AI can take up this aspect to send out auto reminders to both the Auditor, Auditor supervisor, Auditee team and interested parties Support needed (to the Auditor and Auditee team) etc. If an AI agent is created it can throw responses from the repository of FAQs which can be very helpful to both the Auditor and the Auditee team Post Audit Audit outcome discussion between Auditor and Auditee team - No AI here - this needs to be done between the two parties in-person Publish report with timelines to mitigate non-conformities - AI can take up this part and relieve to pick up another audit Seek commitment and send reminders as per the agreed timelines to the Auditee team on timely closures/ mitigation of gaps identified - AI can send reminders to interested parties and keep a close track until all non-conformities are mitigated Support/ consultation needed to Auditee team post the - If an AI agent is created it can throw responses from the repository of FAQs, templates and formats along with some successful case-studies which can be very helpful to the Auditee team to mitigate the opportunities and gaps.