Deep Dave

Hyperautomation can boost RPA capabilities by integrating advance technologies like Machine Learning (ML), Artificial Intelligence (AI), Big Data, Process Mining and Natural Language Processing (NLP) with RPA systems. Let's see how hyperautomation can add capabilities to RPA: 1. Enhancement In Data Processing: The technologies like NLP and Intelligent Document Processing (IDP) makes data extraction and information analysis easier specifically when dealing with unstructured data like PDFs and emails. 2. Intelligent Automation: The technologies like ML and AI will increase system capabilities with data analytics, predictions and complex problem solving. 3. End to End Process Automation: With hyperautomation there is possibility of automating entire business process and not just the individual tasks, thus allowing for comprehensive automation solution. 4. Scalability and Flexibility: Hyperautomation can help with automation across different departments and thus provides flexibility and scalability in automation. 5. Advanced Data Analytics: Integrating RPA with Big Data and Process Mining can help with uncovering hidden data patterns and thus can enhance advanced data analytics through making predictions easier by understanding underlying patterns. Let's talk about industries most likely to benefit from hyperautomation: 1. Manufacturing: Through improvement in PQCDSM parameters (Productivity, Quality, Cost, Delivery, Safety & Morale through RPA enabled with hyperautomation. 2. Healthcare: Through automation in patient data processing, health parameters monitoring and improving patient care. 3. E-Commerce: Through streamlining inventory management, customer service & order processing. 4. Banking & Finance: Through enhancement in compliance, customer service & customer experience improvement and fraud detection. 5. Insurance: Through improvement in claim processing, risk management and decision making. Thus, Hyperautomation will boost RPA capabilities which will have widespread application across almost all the sectors in the economy.

Parametric Analysis is statistical analysis done on known distributions, where we can make inference on population parameters like mean and standard deviation based on sample statistic. Exp. Normal Distribution. There are some key features of Parametric Analysis: 1. Assumption: It follows specific know distribution. 2. Efficiency: If efficiency holds true, it provides precise results. 3. Rely on Statistical theories and formulas. Usage Across Industries: 1. Pharmaceuticals & Healthcare Industries: In pharma & healthcare for new drug development we need clinical trial and bio equivalence study which heavily relies on parametric analysis as it requires inference based on sample data provided that the data follows known data distribution. 2. Manufacturing & Quality Control: For developing robust product we need to ensure that process parameters have high sigma level. Generally, we try to increase sigma level of Critical Quality Attributes (CQAs) through optimising Critical Process Parameters (CPPs). 3. Aerospace & Automotive: High reliability is a must requirement for aerospace & automative industries which requires high precision and accuracy which means it has high application of parametric analysis. 4. Service Industries: In service industry, the arrival rate, service rate and queuing follows specific statistical distributions that also requires parametric analysis. Advantages of Parametric Analysis: 1. High Precision: In manufacturing, if we want to compare machine’s output to specific standard the parametric t-test is the best option. 2. Power of Test: In pharma, while conducting clinical trials the parametric tests like ANOVA can help with checking efficiency of new drug. 3. Wide Applications: Most real life applications follow normal distribution which is helpful for regression modelling (Exp- Energy Prediction Modeling) 4. Simple Interpretation: Process capability analysis through Cp & Cpk helps with summarising how the process is performing against customer defined specification limits. Exp- Moisture content in drug can be measured in Cp & Cpk and can be interpreted on how it performs against specification limits.

In real life scenarios we witness many such failures in a Tech Product or Tech Tool where root cause can be linked to a flaw in the Business Requirement Document (BRD). BRD is a document that articulates product or tool requirement to solve business problem or improve business performance and this document is taken as a guidebook by supplier to develop Tech product or Tech tool. BRD generally consists of the feature, functionality, performance expectation of the product and also the constraints of the project. Thus, it's clear that any ambiguity or inaccuracy may result in potential failure in product or tool. There can be multiple reasons for this: 1. Lack of Articulation: Articulating the exact requirement in BRD is a tough task. Generally, the user has dream Tech product conceptualization in mind but articulating that in exactly the same way can play significant role in success or failure of Tech product or Tech tool. Example: Apple's BRD on Maps could not convey the requirement of precise location with accuracy that resulted in significant map errors, improper navigations and low quality satellite images. 2. Lack of Prioritization: In many cases people want it all in BRD and give all the possible features in Tech product or Tech tool which may result in lack of prioritization to unique and most valuable features. Example: Windows Vista is a perfect example for this where BRD listed so many features without clear prioritization on performance focus. Ultimately, it failed due to frequent crashes and compatibility problems and had to be replaced with Windows 7. 3. Misinterpretation of Business Needs: In many BRDs specificity is missing which may create misinterpretation. Example: If requirement is on support with multiple payment options for e-business application, then developer might miss one of the important integrations (exp-Bit Coin payment option) which may result in loss of potential sales. 4. Poor Definition of Success Criteria: This, I believe is the most important reason as Success Criteria for the Tech product or Tool must be defined with respect to performance, quality, durability and cost effectiveness. Example: JioChat where despite so many Jio Users the app could not succussed as it was lacking success criteria like target for user adoption, engagement and market spread. Thus, it is important to articulate 360deg angle on Business Requirement in BRD so there is zero misalignment.

Decision Intelligence (DI) refers to discipline which combines data science, artificial intelligence and managerial science to make data-driven and the most logical human like decision making. This requires study of advanced analytics, machine learning and use of some cognitive tools to make much more informed decisions. If we consider 4 types of analytics: Descriptive, Diagnostics, Predictive and Prescriptive analytics then the Prescriptive Analytics contributes the most to Decision Intelligence. Since I am from manufacturing industry, let me give the most populist example when it comes to 4 types of analytics: "The Bearing Failure Example" Descriptive Analytics (Reactive in Nature): Here, we describe or summarize historical data to understand what has happened. In Industries, we generally create a metric like "Mean Time Between Failure (MTBF)" to create a baseline on this and try to increase that as high as possible through Kaizens to reduce bearing failures. In modern days, we use tools like vibration testing or bearing temperature measurement with the help of thermostat. In descriptive analytics, we would measure what is average operating temperature & vibration frequency and what is standard deviation for the same. That's descriptive analytics, we describe central tendency or variation of a metric to measure & try to improve upon that. Diagnostics Analytics (Reactive in Nature): Here, as the name suggests we diagnose on bearing failure either with Why-Why Analysis or Fishbone Diagram or we use data analytics like assign the root cause to anomalies or sudden spikes in temperature or vibration. With this, we can investigate possible root causes leading to the failure and take corrective and preventive actions (CAPA) in such a manner that the issuer never repeats. With diagnostic analytics, we may reach to many probable root causes like inadequate lubrication, misalignment or overloaded bearing and take suitable CAPA actions on that. Predictive Analytics (Proactive in Nature): Here, we try to predict bearing failure by applying machine learning algorithms on temperature or vibration with the help of which we can predict an impending failure and once we are confident on the regression model or ML algorithm, we schedule bearing replacement accordingly. In industries, we use Maximo systems for CMMS (Computerized Maintenance Management System) by IBM in which we there is measurement of predictive parameters like temperature and vibration and there is auto maintenance work order created based on alert set on temperature and vibration triggers. Prescriptive Analytics (Proactive in Nature): Just like doctors prescribe the medicines, these analytics expect systems not only to predict but also prescribe best course of action. Like in the bearing example, based on temperature and vibration predictions, prescriptive analytics can recommend the actions like optimized lubrication quantity and schedule, reduction in load or schedule on when to replace the bearing. Thus, prescriptive analytics is highly correlated with Decision Intelligence (DI).

No matter how advanced today's Large Language Models (LLMs) are, there are many limitations due to which LLMs are not ideal to solve all kind of problems. Let's review some of the problem categories with examples which LLMs cannot solve: 1. Problems Requiring Reasoning & Inference: LLMs depend on matching the pattern and lack the deep understanding like human brain. So, in many problems where in-depth understanding and logical reasoning are required, LLMs fail to solve the problem. Example: Apple did one study, where it demonstrated failure of LLMs in standard mathematical problem where in by changing names and numbers led to drop in performance up to 9.2% and when irrelevant details were included, there was drastic drop to 65.7%. 2. Problems Requiring Maintaining Context in Long Conversations: Many a times, we need to give a long context to LLMs for getting a response considering all the constrains and entire context. But they most of the time loose the context in a long and extended conversations. Example: In ChatGPT (If you are extensively using it, then you know!) when we give long context, many a times it forgets earlier context then we again need to reiterate to consider that and reshare the response. 3. Problems With Ambiguity and Incomplete Information: If there is incomplete or ambiguous inputs are given then LLMs may generate incorrect or nonsensical responses. Example: If you are typing open ended question like "What to do if my car is not starting?", the LLM response might suggest checking the battery, but it may also suggest unrelated actions like checking the oil due to no situational understanding. Here, Situational Understanding becomes important as it eliminates many probable root causes in the problem. 4. Problems Requiring UpToDate Knowledge: LLMs are generally good at solving problems using historical data & context till which model is trained, but there is no real time updating capabilities. This means LLMs might give some outdated information. Example: LLM trained before 2024 will not be aware of latest events, leading to inaccurate responses. 5. Biased or False Information: LLMs has large backend data upon which it is developed. If biased data is used in the backend, then LLMs will generate biased or false information which is dangerous. Example: This is recently in news that the countries developing LLMs feed the back-end data as per country's stand on controversial topics (e.g. Land, Wars etc.). Two countries may have different LLMs giving different response to same question as who won the historical war between A & B?

Retrieval-Augmented Generation (RAG) and fine-tuning for LLM-Powered Agent (Large Language Models) are both types of AI language generation methods of Natural Language Processing (NLP) with some fundamental difference in response generation. Let's understand: In RAG, the response generation is based on external source of knowledge that helps with real-time information update & the same information is used to augment the model's response. Basically, instead of relying on internal knowledge used for model training, RAG has kind of "Open-Book" approach where any additional or real-time information is looked from external sources. Whereas in Fine-tuning for an LLM-Powered Agent, there is pre-existing internal data source which is fine-tuned through continuous improvement on a specific set of data to specialize it for specific task or domain. Say for example, we are adding specific books in internal data sources for generation of more advanced & niche responses which industry or domain specific. Application for RAG: Dynamic & real-time knowledge requirements (e.g. stock price, latest news, latest research etc.) Model Size Reduction (as no need for huge internal data source required) Cost & Efficiency (requirement of low or no internal data eliminates requirement of high energy consuming servers for storing internal database) Application for Fine-Tuning with LLM Powered Agent: Nuances & Specialized Tasks (e.g. application with special domain language, knowledge and jargons like medical, legal etc.) Consistent Tone and Style (nuanced in line with ask in the prompt) Hybrid Use Case (RAG & Fine-Tuning with LLM Powered-Agent) In cases where we need the response that is specialized in language and also dynamic with respect to real-time or latest information update, we need both internal and external sources for response generation. In this scenario, we need to explore the Hybrid model. Examples: Let's say we have created Financial Advisor Chatbot which uses finance related language and tax laws using (Fine-Tuning with LLM) as well as gives advice considering latest market dynamics (stocks, bonds, investment avenues etc. - RAG). In this case, it's best to use both internal and external sources to generate ideal response. To summarize, hybrid model is ideal scenario for specialized and real-time response generation.

Named Entity Recognition (NER) - A technology used in the Natural Language Processing (NLP) to identify and classify entities in text. Say for example, we are saying that “Narendra Modi was born in Vadnagar” then NER should identify “Narendra Modi” as a person & “Vadnagar” as a location. Above was a simple example but let’s add some ambiguity. Let’s say we are typing in that “Apple manufactures iPhone”. Here, NER should be able to identify “Apple” as name of the organization and not as fruit. Hence, NEP tools should have capability to identify entities like names of people, places, organizations, dates in right context even after ambiguity as mentioned in above example. Now, logically thinking with human brain we can distinguish Apple as a fruit or Apple as a mobile manufacturing company by looking at the context in which the word is used. Let’s see how NER systems in Natural Language Processing deal and process when ambiguity: 1. Through Contextual Analysis: In GPT-4 the NER technology uses contextual analysis using pre-trained language models like BERT or GPT-3 through which context is understood and accuracy is improved. 2. Google Bard: Data augmentation & context-aware models. 3. LLaMA: Heuristic rules 4. Claude: Contextual embeddings, attention mechanisms, multi-task learning and knowledge graph To summarize, the basis for all LLMs is contextual analysis, pre-trained models, heuristic rules & attention mechanisms.

Replication followed by benchmarking is one of the most effective and the easiest methods for lean six sigma project identification. The greatest advantage here is that we get to replicate the projects with proven success and results. However, if we seek replication of patented products or processes then we may face legal challenge due to violation or infringement of laws related to Intellectual Property Rights (IPR). Let's see below the major challenges and strategies or approaches to overcome these challenges: Major Challenges in Replication of Patented Products (DFSS) or Processes (Six Sigma): 1. The Legal Aspect: The IPR laws limit us to replicate the design or technology used behind a patented product or process. Failing to comply this our company may face lawsuits and loss of reputation. 2. The Innovation Aspect: Companies that heavily rely on replication may miss the huge opportunities of inhouse innovation. Different companies have different goals and strategies, in these scenarios replicating technology or process patents dealing with specific issues or goal of other company may not be relevant to us. Instead, the companies should train people and start an innovation drive of their own. 3.The Information Aspect: Generally, the detailed information on "How?" part is missing behind patented products or processes. In this case replicating the exact concept is a challenge. Strategies & Approaches to Overcome Challenges on Replication of Patented Products (DFSS) or Processes (Six Sigma): 1. Adaptation of Principles Instead of Process or Technology: Instead of replicating exact process or product design, can we think of replicating principles behind? This helps with avoiding legal risk as well as capturing the essence of patent technology. 2. Collaborate with Patent Filing Organization: In the cases where it is difficult to replicate, we can think of collaborating with patent filing organization through licensing agreement. This is the best approach when we do not have internal capability to replicate the exact patent technology. 3. There are Multiple Solutions to a Single Problem: While we know this fact, we can train & encourage inhouse team to innovate different solutions to the same problem. 4. Innovation In Six Sigma: In Six Sigma Projects, typically in Improve phase when we explore different improvement initiatives, bringing in innovative thinking and methodologies (exp. SCAMPER) can help the teams identify innovative solutions. This way we can avoid necessity to replication to patent technology and also get the end result. To summarize, we have strategies like collaboration, adaptation, innovation, inhouse R&D and capabilities development to deal with the challenges.

Standard Work and Work Instructions are generally more prevalent in manufacturing set-up. “Standard Work is a structured approach where optimal sequence of steps/work and time is defined around work routine to minimize variability and maintain efficiency.” Vs “Work Instructions are generally far more detailed, providing specific step wise instructions about how to carry out the individual task within the broader concept of standard work.” Difference Table: Standard Work Work Instructions Covers the overall process. Focuses on specific task, providing detailed steps for the task. High level, summarizing key steps and timing. Detailed, including step-by-step instructions. Designed to minimize variability and maintain efficiency. Designed to ensure a task is done correctly, consistently, and according to standards. Focus on optimizing the flow of the process. Focus on individual task execution. Typical audience include supervisors, managers, and operators. Primarily for operators or workers performing the tasks. Focus on workflow. Focus on visuals like diagrams, photos, or flow charts for clarity. Which is more important? IT DEPENDS!! Standard Work is more important for supervisors & managers for ensuring all priorities are met through daily, weekly, and monthly defined tasks. Standard Work should also be dynamic in the sense that standard flow of the process should continuously improve in line with changing company priorities. Whereas Work Instructions are more important at operator or worker level where following step-by-step instructions is crucial for ensuring quality right first time. Thus, rather than viewing one as more important than the other, it is their synergy that truly drives operational excellence.

Let’s start with what Gamification is! Have you ever tried this? May be avid book readers like me will relate. “Imagine you are reading a book, and the books is really big. You find it boring the halfway BUT you want to brag about completing the book to your book reading community…so you have created one technique that every 10 pages you complete, you will reward yourself. I used to take peanuts jar with me (some affluent people may keep the jar of almonds or cashew nuts!). The point here is that this technique helped me. It was like a carrot approach where upon finishing 10 pages, I would reward myself some peanuts…. For me this is the best example I can relate with Gamification. (For reference book was Don Quixote by Miguel de Cervantes) “Gamification is creating a game-like situation for making routine, boring or tedious activities to make them more fun, engaging, and enjoyable.” If we consider learning with Gamification, then it would be one of the best learning techniques as our brains always works well with dopamine release every time you achieve milestone in a game be it a videogame, PlayStation or Gamification of actual real-life activity. Gamification involves different elements like... · Avatars · Simulation · Storytelling · Mystery Solving · Point Systems · Competition · Badges · Leaderboards Also, some of the new age technologies like metaverse and digital twin technology are based on core philosophy of Gamification only. Gamification to improve LSS training effectiveness (With Examples): · Avatars: Imagine an online training program where participants create Avatars and upon completing trainings, completing assignments or projects the participants can change costumes & ranks. This will encourage participants complete trainings and assignment with more enthusiasm. · Simulation: Imagine if participants are given a business problem in a simulated manner where they are given step by step story and context for each DMAIC phase such that they get a feel that they are solving actual business problem. · Storytelling: LSS concepts are always easy to understand with storytelling. when I think of median, it’s always easy to understand with an example of “Average salary of 5 people in the room and what happens to average if Bill Gates enters the room?” · Mystery Solving: Defining mystery-like scenarios that helps trainees develop critical thinking skills, encouraging them to creatively apply Six Sigma methodologies to uncover hidden inefficiencies or defects in processes. · Points Systems, Competition, Badges & Leaderboards: Imagine a point system for participants to promote healthy competition where every time based on training completion, tests, problem-solving and projects they get points and upon completing milestones (e.g. 1000 points, 2000 points.) they get badges which is like sense of pride for them. Here, creating leaderboards will encourage participants to put more efforts to top the Leaderboard (Just like with this answer I am hoping to be the topper in Benchmark Forum Leaderboard!)

Post-Purchase Rationalization occurs when customers convince themselves that their purchase decision was the right one, even if they are not completely satisfied. This can distort customer satisfaction metrics / KANO model in Lean Six Sigma projects, leading to misleading conclusions about a product or service's quality. For instance, consider a scenario where a company introduces a new product, and customers who spent a significant amount on it might report being satisfied to avoid feeling regret, even if the product didn't fully meet his expectations. This can result in misleading satisfaction scores, preventing the identification of genuine improvement needs. Analytical Methods to Identify and Rectify the Impact: 1. Behavioral Analysis: Off course in above cases customer may not repeat purchase, so tracking a metric of "repeat purchase" will give more insights. In addition, we can track metrics like return rates and complaint frequencies. 2. Drafting Better Customer Feedback Survey Questions: Sometimes survey is designed in such a way that we get the best CSAT score, instead we can design precise surveys asking about specific aspects or features of the product. 3. Surveys At Right Time: May be not at the time of purchase but if we take survey somewhat later during product lifecycle where customer has long exposure of products or services then may be real CSAT score will come out with actual concerns. 4. Net Promoter Score (NPS): May be at the time of purchase the customer gives good CSAT score due to this phenomenon, but in this case if customer is actually not satisfied, he will not recommend this product or service to others, so yes NPS can be another metric to check! 5. Comparative Analysis with Competitors: Benchmarking with features provided by other best product/service providers may help offset this phenomenon. By using these methods, Lean Six Sigma teams can better detect and address the effects of post-purchase rationalization.