Hamid

A Prompt + Flow-based AI solution can surface hidden cross-process patterns by following a process: 1. For starters it will need to analyze the inputs which it can summarize issues by topic across the various departments (HR, Finance, Operations, Tech Support). 2. Thereafter it can further categorize the issues (e.g. Process, Technology, People) or add metadata that can be relevant to the context such dept., location, team, dates etc. 3. The next step in the process will be to detect pattern using NLP by analyzing the text and identifying recurring themes, keywords and phrases. AI will then further group the observations and revealing patterns and note any correlations across departments. 4. The Root cause analysis can then be used by the leadership as actionable insights and the AI can even help with suggesting projects and initiatives, corrective and preventative actions. 5. The final output will be to generate insights and present this back visually via various dashboards and graph formats to the audience. Further to this answer to clarify the inputs the AI may need: Brief descriptions of the issues and observations from the different departments. If available, any possible metadata too (e.g. dept. location, team, dates etc.) will be helpful as mentioned above which it can use as part of the correlation exercise as well as an output when reporting on the common issues.

I genuinely do believe all front and back-office enquiry handling in the BPO industry can and will be handled by AI. For the purpose of this question, I would like to focus on Customer enquiry handling. The problem statement will have to be long wait times; the enquiry process itself is time consuming and labour intensive. With agents dealing with high volume of repetitive tasks/enquiries the margin for error is always fluctuating depending on time of day, energy levels and variables that are not consistent which is inherent of a human being. Further to this, these elements then lead to poor satisfaction levels due to the above-mentioned challenges when using only humans for these types of enquiries. The motivation would be a saving in operational costs and less waiting and handling times for customers. Also, agents can be redeployed to more meaningful tasks and the AI can deal with the mundane repetitive tasks that often contribute to lower agent morale, decreased morale and eventually burnout. With a reduced error rate and an ever spunky, energetic AI agent, the Customer Satisfaction performance can only be impacted positively. The appropriate approach would be to employ the Fine Tuned LLM with flow + prompt-based design and here’s why: This approach fits best because it: 1. Will utilize existing LLM capabilities and adapt them to customer enquiry handling. 2. As noted in the BPO industry we deal with high volume transactional, repetitive enquires, this approach provides the flexibility and scalability to handle these volumes. 3. Because this approach will also allow for personalized responses and is quite efficient, it will definitely lead to improved Customer Satisfaction. By employing the fine-tuning an LLM with a flow + prompt-based design approach, BPO's can automate customer interactions/enquiry handling, reduce operational costs, and ultimately improve customer satisfaction.

When researching these approaches, one needs to consider a few factors like complexity, data, performance, and time to develop and deploy. Each approach has its own strength and weaknesses, and one need to be discerning when reviewing the goals of the AI project. Below are some advantages and limitations per approach. Further below is a table with some considerations and a quick view to determine which approach would be suitable for your potential use case. AI Solution Approaches: Comparison and Contrast 1. Conventional AI Models and Methods - Advantages: -. This method is generally understood and well established. - Uses less CPU power generally. - Limitations: - May not handle complex tasks. - Requires lots of specific expertise. 2. Fine-Tuning Existing LLMs - Advantages: - Uses existing knowledge and current capabilities. - Quicker to deploy. - Able to deliver a high performance. - Limitations: - Performance can be inconsistent. 3. Training a New AI Model from Scratch -Advantages: - Can be set up to specific needs. - Can be better in performance. - Limitations: - Resources can be expensive based on CPU power requirements. - Needs large datasets. 4. Designing Solutions with Flow and Prompt Engineering - Advantages: - To develop this can be done quick and updated too. - Can make use of LLMs and avert high costs. - Limitations: - This needs higher expertise in prompt engineering.

This is a very interesting question as I have noted based on some basic research and following the AI technology, trends and applications that almost all functions within the BPO industry can be replaced or automated by AI with minimal human reliance review protocols for exceptions and where more human based decisions may be required. I believe AI can and will replace all front office and back-office functions reducing dependency on human agents. These functions will include providing services like Customer Support/Customer Service/Customer Care which is generic in nature and spans across all aspects of the customer's in life (end to end) journey from Onboarding, in life, Billing, Service interruptions, upselling, additional services products to offboarding however there are key elements that the industry need to guard against which could include the below: 1. Complaints that will impact brand and reputation, Complaints that will be escalated to Industry ombudsman and regulating authorities that can or may result in heavy penalties and potentially reputational damage on comparative websites e.g. Trust Pilot which has built in league table rankings per industry. 2. Where AI hallucination could result in losing customer base e.g. Collections process where customers specific circumstances need to be considered and does not necessarily fit into a specific pre-determined description or bucket. An example of this is a customer that has been incapacitated and need to share proof/documentation of this, and a special exception will need to be applied to allow this customer special consideration instead of deactivating/disconnecting/pausing services/de-energizing energy supplies which can have far reaching negative consequences for the customer and the organization. Ultimately, I still do believe a large percentage of BPO industry internal and external services can be automated via AI.

Currently we are always seeking opportunities to optimize the operations and leveraging support functions to ensure the end client receives an enhanced or elevated customer experience. The way we do this is by deploying process improvement projects and initiatives based on Customer Experience Survey Scores, survey responses and subsequent Root cause Analysis (all completed by humans). So the “as is” steps are: 1. Trigger CX surveys to customers post a closure of a query on the CRM. 2. Receive survey responses and have analysts do RCA based on a predetermined set of rules or criteria. 3. CX Insights manager then reviews the outputs of the RCA and ranks it highest to lowest in terms of impact to customer. 4. Once this is refined and documented, the Operations leads will connect the client for alignment and approval on agreed projects and initiatives that will potentially improve CX. My “To be – AI integrated approach” would be as follows: 1. (Automation): Trigger CX surveys to customers post a closure of a query on the CRM. 2. (AI): Employ an AI bot to handle received surveys and generate Dash views based on customer sentiment that are curated by means of logic and criteria which informs management of the CX impact and articulate issues in a way that it is ranked by priority highest to lowest. 3. (AI): Also rank dash views by order of positive impact if resolved, so show forecasted scores if top 5 issues are fixed. 4. (AI): the AI agent to share suggestions of approach to be applied and share to operations and client. 5. (AI): Use AI agent to track progress of progress and initiatives activities. 6. (AI): Use AI to track progress/trend of scores based on impact of activities. 7. (AI): Use AI to track progress/trend of scores based on impact of activities vs failure points where no projects and initiatives were actioned.

Where the AI agent must balance 2 or more competing objectives, accurate and helpful whilst ensuring the customer feels supported and valued requires 3 elements to be in place: Logic; Rules and Signals. One such scenario would be when employing a Customer Support AI Chat Bot. To guide the AI agent, we can use a combination of logic, rules, and signals. Here's a possible approach: Logic: 1. Weighted scoring: Assign weights to different aspects of the response, such as accuracy, completeness, and relevance. The weights can be adjusted based on the customer's preferences and priorities. 2. Threshold-based evaluation: Set thresholds for response time and customer satisfaction. If the response time exceeds the threshold, the agent can adjust its response to prioritize speed over accuracy or completeness. 3. Contextual understanding: Use natural language processing (NLP) and machine learning algorithms to understand the customer's context, tone, and language. This can help the agent tailor its response to the customer's needs and preferences. Rules: 1.*Response time tiers: Establish response time tiers, such as: - Tier 1: Quick responses (e.g., < 1 minute) for simple queries. - Tier 2: Standard responses (e.g., 1-5 minutes) for more complex queries. - Tier 3: In-depth responses (e.g., > 5 minutes) for critical or complex issues. 2. Customer segmentation: Segment customers based on their preferences, behavior, and value to the organization. For example: - VIP customers may require more personalized and rapid responses. - High-value customers may require more comprehensive and accurate responses. Signals: 1. Customer feedback: Collect feedback from customers on their satisfaction with the responses. This can help the agent adjust its weights, thresholds, and rules to better meet customer needs. 2. Agent performance metrics: Track metrics such as response time, accuracy, and completeness. This can help identify areas for improvement and optimize the agent's performance. 3. Contextual signals: Use contextual signals, such as the customer's location, device, or previous interactions, to inform the agent's responses and improve customer satisfaction. Weightage: · Accuracy = 0.4 · Completeness = 0.3 · Relevance = 0.2 · Response Time = 0.1

The most relevant and practical scenario that comes to mind is employing an AI Chabot which is capable of handling Customer Support/Service/General enquiries. The flow of the chat bot should be as follows: 1. Direct customers to self-help information and FAQ's. 2. Guide on troubleshooting and 3. Escalate or route to a specialized/complex/complaints dept. operated by humans to further intervene and resolve. One such scenario is when the AI chatbot is dealing with Gas and power (Energy Utilities) customers that have requested an invoice to be rebilled to an accurate read as it constantly bills to an estimated read resulting in an inaccurate invoice. Post all the rules of elimination via troubleshooting and there is still no satisfaction, the AI chatbot will need to route the error/issue together with the interaction and troubleshooting history to a technical dept./Human so that a physical inspection can be conducted on site to establish if there is any physical impairment or issue with the physical meter. Thereafter a physical meter fix/replacement will need to be actioned, and the human agent can then post their updates and fixes to the customer and then hand back to an AI chatbot once the client/end customer has confirmed they are satisfied with the rebilling to an accurate bill/invoice.

Defining a problem statement where it captures the minds and agreement of all stakeholders can be very tricky as all stakeholders have their unique POV of a problem or a goal for that matter. Below are some practical steps that I would suggest: Gather a team from every supporting and operational department. This will ensure inclusivity and representation from all areas creating a diverse and dynamic team. Collect data and insights from all areas specifically related to this issue/problem. Establish inclusivity and participation from all stakeholders/representatives so they feel their POV is heard and acknowledged. Always ensure that communication channels are open and 2 ways, so the team feels valued which is not always an easy feat. This can be facilitated via workshops and create a safe space, so each team member does not feel overshadowed nor insignificant irrespective of the contribution. Clear up any assumptions, misunderstandings and point to the data as an objective source of reference. This will also validate any assumptions and take the team with on a journey of data analysis and review so it can be a collaborative arrival at the end goal, which is defining a factual and data validated, objective project problem statement.