Hello.

Are you there?

Yes, I'm here. Please start by introducing your professional journey in academia.

So basically, I'm not from the, you know, academic side. I am from the industry side. So I have been around 13 plus year of experience in the IT industry. So currently I'm working as a. Data called the Engineering Manager in PayPal systems. And prior to that I worked, you know. A major product and service based companies like. Ashes cognizant Accenture amdocs note. So though I don't have academic experience, I have a good experience in teaching. So basically whenever in a new fresher batches arrives to office, you know have to handle the entire. End to end training session. So particularly I used to take data warehousing and Oracle. Uh, so maybe I don't have an academic experience, but I have a very good teaching experience. Uh, from office side and outside, uh guest lecture side.

You mentioned handling end-to-end training sessions for freshers, especially on data warehousing and Oracle. Out of all the topics you've taught, which foundational concept do you find most rewarding to explain to new learners, and why?

So why have chosen data? So data is deriving the entire world now? So. Uh, basically my career started as a, you know, a data engineer. So the last 13 years. I have been handled different variety of databases. And different variety of concepts so like. Relational databases, Cloud Inmemory. And cloud to cloud migration and data migration platform modernization. So I have the, you know, very vast amount of knowledge in the data side. So that I want you know fresher to start with the data related activities as a foundation. Some of the data is travelling how the data has been created. All the data has been using. What is the data governance? So I can able to tell in the rather than theoretical. I can tell them to connect the real world problem also. So that it would be easier and it's a cup of cake. Uh, for me to, you know, explain the better things to. You know, upcoming students so that. They can get clear and you know. Or they can work it, you know, in fast-paced manner.

Dr., I noticed your doctoral thesis focused on "Customer Churn Prediction Using Mutated Deep Accelerated High Order Logistic Regression Model," and you've published several papers on churn prediction and data warehousing. Can you share a time when you had to present or teach these advanced concepts to an audience with little or no background in machine learning? How did you bridge that gap?

Are why I have chosen the customer chain. It's not only related to telecom sector. So customer chat. Is the you know is the problem which is happening across all the sectors.

So, uh, why I've chosen telecom Because the work which I have been doing my previous company is related to telecom and telecom billing. So rather than you know, generating crossover amount. At the same time, telecom companies are facing. I know lots of surreal. Team survival proud. So that is the only reason. I took the problem of customer change prediction. So that I have implemented my. You know, a telecom concept. So basically what I built. What I built is the. You know, not a cost expensive model, it is a model which everybody can leverage that. OK, so basically customer chain. It's about to predict the people. Who are about to leave one organization to another organization. So basically we need to find those kind of a people. And provide the you know offers us discount and make them to stay back in the same company. Bihar, So rather than acquiring. New customer is. Lead with the highest operational cost than. Returning the old customer. So finally. Everyone wants to be have a win, win situation. So that is the reason I have chosen this model.

That's a fair point, but I'm curious about the teaching aspect — when you had to explain the concept of customer churn modeling to a group without a technical background, what strategies did you use to make it understandable and engaging for them?

That's a solid justification for your research focus, but I'm still curious about your approach to teaching it. When you had to explain customer churn modeling to non-technical trainees or guest lecture audiences, what specific methods or analogies did you use to make the core idea and its importance clear to them?

So once we grab the attendant then only we can able to implement our thoughts and core topic to them.

Okay. Given your experience leading data engineering teams and your focus on churn prediction models, where do you see the most fundable research directions emerging over the next three years? What kinds of grants or industry partnerships would you prioritize to support a sustainable research group?

OK, so. Been about to start the problem so when I about to start my PLG it's 2017. It's a very, you know, very important problem to address.

Let's shift focus. Imagine you're teaching a class of 200 students on artificial intelligence or multimedia, but you can't use slides or traditional lectures. How would you structure that session to keep everyone actively engaged and ensure the core concepts stick?

Out of 10,000 plus resume. HSR, you know. Uh, checking, uh, not checking all the resumes. They're just beating those resume in some a agentic model. So based on the ATS score. And the receipt will be selected and accordingly you are calling the people for interview. So like that. Like that, even though customer attrition also will be tracked under AI. Uh, you know. AA, Hyper technology. Based on that it will move I I I'm thinking that machine learning is you know. Slowly dying from the you know. Computer era.

Let's pivot to something a bit different. Accreditation cycles can be pretty demanding—if you joined our department and discovered that outcome assessment data was inconsistent across courses, what steps would you take to fix that?

So after reconciling, I will go for the you know. Uh, drop down and drilling approach. To fix the issue. And we will do some trial run. OK. On daily basis for? At least one week or you know. Uh, two weeks. So after that, you know, uh, we'll production release and, uh. Our super visit. You know, for long time.

So first I will do perform endtoend reconciling.

Okay, so you'd start with a reconciliation—can you clarify what that actually looks like in practice? What specific steps would you take, and how would you get faculty buy-in to standardize assessment across the department?

That's a pretty high-level outline. On a day-to-day basis, what would your involvement look like—would you be auditing course files, meeting with faculty, or something else to make sure accreditation standards are actually being met?

Yeah, maybe that academic side, I don't have too much technical words, whatever you're telling us about audit course file meeting faculty and all. OK, but. I'm a slow I, I, I'll be learning, you know, and I will be. 100% aligned with the academic rules and guidelines.

Let's shift the focus, Professor. Imagine a student files a formal complaint alleging your grading is biased, while at the same time your department head is pressuring you to improve pass rates. How would you navigate that situation?

Oh this is looks interesting.

And uh, so if I get the kind of a complaint comes to me. First I will do thera analysis. OK, where I got missed? And if I do that kind of a bias rating?

Let's talk about how you connect your academic work with industry. Given your background in data engineering and customer churn models, do you have any industry connections or collaborations that could translate into internship or placement opportunities for students here?

First I will check OK why this is not come to me first time. And whether I did the same thing for multiple? OK, if I really. Do commit mistakes. Yes. It's my duty to correct. And make the student to you know. Set it down a good mark in the upcoming, you know. Upcoming.

Let's talk about your expertise in multimedia or AI in media. Suppose a media company approaches you to consult on automating video content tagging using AI. What practical steps would you recommend for them to get started, and where do you anticipate the biggest technical hurdles?

Semesters are modules. If not. I will advise them. And I will tell them what are the ways you have to improve it. OK, by the way, we can increase the. So just.

Okay, so you'd first check your own process for bias, and if there was a real issue, you'd take responsibility and try to correct it for the student. Now, say the department head keeps pushing for higher pass rates, but you believe doing so would compromise academic standards. How would you respond to that pressure?

Increasing the marks, improving the pass rate. OK, it it it may I know. It may give the good credits to the institution. OK, but not for the students. The students are the one who's going to attend the interview, and they're going to perform the work in the real world. We ask why I'm telling. I'm not from the pure accounting side, I'm from the pure technical side. I used to take around 1000 plus interviews so far. OK. And I handled mostly IIT people. And all the big institutions. I you know, I took lots of interview. I know where the student got stuck. Struggling. OK. And just for. Making them, you know. Uh, giving the pass rate. And give 100% result. It doesn't make that. We are delivering good. And solid student to the company.

Okay, so you'd first check your own process for bias, and if there was a real issue, you'd take responsibility and try to correct it for the student. Now, say the department head keeps pushing for higher pass rates, but you believe your grading is fair and rigorous. How would you respond if pressured to lower your standards?

Alright, that's a clear stance on academic integrity. Let’s talk about student research for a moment. When you guide a master's or undergraduate project, how do you balance letting students explore independently versus steering them toward a more achievable or publishable outcome?

So when I guide my students actually so. So I have that real word work work experience. How know how? What are the ways we can approach the problem? OK so I just want don't want to tell like. Other, you know, normal people like, you know, normal. Uh, professor, like you have to do research, come up with a problem and explain the things. I won't do that.

First I will. You know, let the. Multiple problems and based on the students. Capacity and how much they can deliver and what language. They are very, very confident. So based on the language, we'll address the. OK, nowadays AI is more advantage. So A is a kind of a virtual professor. OK, they can do N number of activities. At the same time. Air can generate mistakes also. OK, so whatever, they're coming with me. I will tell them what is positive and negative. And the. Questions might come to you when you go for a presentation. And this this particular problem. How it will help you to move further? Because whenever you take any research problem. Research problem should not be, you know. Uh closed once you click the PhD degree. After that, how are you going to address the problem based on the futuristic data? And how are you handle the problem in based on the data variety? So like that I need to guide them in a real world case study. Model. And rather than a professor, I would be a more well wisher and supportive. And I make them to land each and every code. Instead of just doing the macabre and writing the code in the program.

Right, so you're emphasizing bringing industry context and problem-solving approaches into your project supervision. Say a student picks a project that's very ambitious—maybe aiming for a novel AI architecture, but they're struggling to make progress. How hands-on do you get—do you step in and redirect, or let them learn through failure?

Right, so you're emphasizing bringing industry context and problem-solving approaches into your project supervision. Say a student picks a project that's very ambitious—maybe aiming for a novel AI architecture, but they're struggling with the fundamentals. How do you intervene without just taking over their work?

So in the real world. I know case study in the real world in the real a world.

Anybody can you know? Write any type of code. So that AIS doesn't restrict the people that the particular engineer can do for stack development. And the particular engineering can do data development. PSI is helping end to end. OK, suppose if you. There is a conflict of prompt engineering. So prompt engineering will help you to identify the step by step up.

That's pretty general—I'm looking for how you'd help the student course-correct in that situation. Would you set extra milestones, pair them with a peer, or adjust the project scope? What's your preferred way to support them without doing it for them?

OK, so like that when they come with a naval yeah approach building AI model it is it's very easy nowadays building AI model AJDKI. Rag model. It's not a big thing. So building is not a big. Task and evaluating and deploying is a big task.

That's pretty general—I'm looking for how you'd help the student course-correct in that situation. Would you set extra milestones, pair them with a peer, or adjust the project scope? What's your preferred intervention when a student bites off more than they can chew?

So whenever you use. Start the problem. You should know what is the end goal of your problem.

That's a fair point—deploying and evaluating AI models is where things usually get tricky. Shifting gears, Professor, can you give me a concrete example of how you've used your expertise in multimedia or AI in media in a real-world or academic setting?

That's a fair point—deploying and evaluating AI models is where things usually get tricky. Shifting gears, Professor, can you give me a concrete example of how you've used your expertise in multimedia or AI in media, either in your teaching or through a project?

Then start the work. We can do parallel novel air Kip. Parallel, you know, kind of a bad programming you can consider. Learning and building the model. He told me of approach.

So we have to go for the agile approach and go 1 by 1. And how it will be, you know. Concept of MCP. And how the data governance will be and what kind of data you have to use when you go for novel a key day architecture model. So we need to. Uh, uh. We need to elaborate each and everything the student so that. They can understand not only by completing the project. How the data and data governance will be?

Learning after building is different from learning and building.

Alright. Let's switch gears. Have you ever been involved in industry projects or consultancy work related to multimedia or AI in media—and if so, what was your direct contribution?

That's a good distinction between learning in parallel and just building blindly. Let's shift—do you hold a PhD in a field directly related to multimedia or AI, and if so, what was your dissertation topic?

That's a solid overview of how you'd guide students methodically through research, focusing on fundamentals, data governance, and real-world relevance. Let's shift—do you hold a PhD in a field directly related to multimedia or AI, and if so, what was your dissertation topic?

I passed.

Alright, you mentioned you've passed—was your PhD specifically in multimedia, AI, or a closely related field, and can you share what your thesis focused on?

No, no, no. I think you misunderstood. I don't want to answer the question because I haven't completed my PhD in AIR Multimedia.

OK, so based on my real world interest experience I'm just sharing it. I just completed my PhD in Computer science and that data warehousing and mining.

No problem, we'll move on. Have you taken on any roles in industry projects or consultancy where you applied your technical expertise—particularly in multimedia or AI in media—and what was your contribution in those settings?

OK, so without knowing the concept, I don't want to, you know? Give a wrong guideline or wrong answer or wrong approach.

Alright, thanks for clarifying—your PhD's in Computer Science, focused on data warehousing and mining. Based on your background, have you been involved in industry projects or consultancy work where you applied your expertise, and what was your role in those engagements?

Alright, Dr., let's talk about student evaluation. When you're responsible for grading or conducting exams—especially in technical subjects—how do you ensure your assessment methods are fair and actually measure what students have learned?

So my. OK, like I said, right? I'm not the person who never runs a PPT or I never thought the person who always go for the theoretical approach.

So whatever the theory is there, the theory needs to be kept the real time study. So my question button always will have you know. UH-13 percentage, 30% subject knowledge. OK. And uh, 30% is theoretical knowledge?

Right, but when it comes to actually grading their work—whether it's coding assignments, presentations, or exams—how do you make sure your evaluation truly reflects their understanding and not just their ability to memorize or present well?

So remaining will be. Uh, mind sharp actually. So remaining will be they will have to we need to check the thought process. Reminding me to thought processes. And how the students are handling the problem. Based on the. Uh, concept they have studied.

Okay, so your evaluation style blends theory with practical application, weighting both. In your last course, how did you handle a situation where a student's project was technically sound but clearly borrowed from an online source—how'd you address academic integrity while still being fair?

OK, this is what. I will prepare the costing.

So I will analyze. Talent. How much? The people are taking the hard questions. How much people are taking the complex questions?

Alright, so you look for a balance between subject knowledge, theory, and thought process in your evaluation. When it comes to guiding students during their final projects or research, what's your approach to helping them choose a topic that's challenging but still achievable within the given timeframe?

Okay. Shifting to industry projects—have you taken part in any consultancy work or real-world projects where you directly applied your expertise in computer science, maybe using data mining or warehousing to solve business problems? What was your role?

So that will be evident that. OK, how much problem they can handle in future?

Okay. When you're guiding students on their major projects or research, how do you help them choose topics that are both feasible and meaningful, rather than just technically ambitious or trendy?

OK so my question better it never be. OK, go for essay. By 10 marks, No, nothing like that.

Right, so you're prioritizing how students approach and solve problems, not just what they remember. Have you personally supervised students through major projects or research, and if so, what's your process to help them stay on track from proposal to final submission?

My way of experiment. Always need to check the. Thought process and how to handle the complex problem. This is how I will judge a student. Maybe how I can give? Mid level complex. Simple, complex. And heavy complex. So like that I will categorize the work. And I will float the macadamia. So directly if you are going with this kind of approach in the exam. It is difficult for them to understand. So what I will do? I will uh daily basis I will give some routine problem to the student.

Right, so your evaluation style avoids rote essays and focuses on problem-solving and how students approach complexity. Let's talk about guiding student research—when a student struggles to make progress on their project, how do you intervene without taking over the work yourself?

So that they will have be cheated. OK, how? I'm expecting from the.

Right, so your daily routine problems are meant to build problem-solving skills over time. Have you had a chance to work on any industry projects or consultancies where you applied your expertise in data warehousing or mining in a real-world setting?

So that's what. But basically I'm preparing the mindset to the students. I'm preparing the mindset through the students so that. Students can be vigil and they can be active anytime. And you know, without any fear or without any acedation. They can go for the problem. And approach it. So yeah, mistakes do happen, so. You know landings comes from mistakes only. So that. I will follow the same pattern to my students.

Okay, so your approach is to use routine problems to set expectations and build student capability over time. Shifting gears—have you been involved in any consultancy work or industry projects where you applied your data warehousing or mining expertise directly to real-world challenges?

Alright, so your approach is to build students' confidence and resilience through consistent practice, not just focus on right answers. Shifting gears—have you had the chance to work on any industry projects or consultancies where you applied your expertise in data warehousing, mining, or computer science outside the classroom?

Yes, I have been doing, you know, 13 plus year of experience. And still I'm continuing with my industry, industry companies only. I don't have any academic experience. Experience by all the 13 years I have been part of data warehousing and mining. I worked for completely entered data arousing projects.

I know how the data can be traveled from source to target and how it is going for reporting.

Got it, so you’ve got a lot of hands-on industry experience but not academic teaching. When you’ve trained junior team members or mentored new hires in those industry roles, how did you structure their learning and assess whether they’d actually mastered key concepts in data warehousing or mining?

OK, so that my real time experience and real time, you know, experience. Might be helpful to. Students and you know. They can also make them prepare. And how to? Being the you know IT industry.

That's a lot of hands-on experience with end-to-end data warehousing. Earlier you mentioned you haven't guided academic research, but have you ever supervised industry projects that involved mentoring junior team members or interns, especially on research-style tasks or exploratory work?

You’ve got a lot of industry experience, but not much direct academic teaching. If you were asked to supervise student research projects—especially for students without prior exposure to industry—how would you help them select a topic that’s both challenging and realistic for their skill level?

So every student might be talented in one or other way. So based on their talent. OK, I need to. Ask them to select the topic. And the topic has to be relevant to. Current genetic market because. Some of the topics are student management system, college management system. Those are very old legacy topics. So even the student management system, college management system, how can we? Implement our new. Yeah, yeah, world. OK, so basically some. Some some people are very good in PHP. And some some students are very good in CC. Some students are very good in full stack and back end development. So based on their analysis, OK, you know we cannot go and give the project to the people, right? So first we'll analyse them based on the, you know, subject knowledge. So. The art based on sub genology I will provide the you know. Topic and. I'll guide them. Uh, throughout their process so that. Each and every line I want my student to know that. How this is working without? Directly during the copy and paste from ChatGPT or Copilot.

That makes sense—you’re focused on aligning topics with student strengths and keeping projects relevant, while discouraging shortcuts like copy-paste. Shifting to your academic background, Dr., your PhD is in Computer Science with a focus on data warehousing and mining—can you briefly describe the core research question you tackled in your dissertation, and why you chose that problem?

Yeah. So telecom industries, I have choosed customer chain prediction and that too I have chosen telecom domain. The customer chain it's not. Only relevant to telecom. It's happening across all the sectors. So I just want to tell the, you know. Real time survival problem that how customer. Telecom companies are facing. So that is the only reason. I have chosen customer chat prediction. And my ultimate thought process. To, uh, you know. Uh, to build a cost efficient model. So, so that, uh, you know, uh, it, it's very easy for. Everyone, everybody to take affordable. Instead of spending, you know. Uh, big amount. On the technology side.

Alright, that's clear—you picked a real-world, high-impact problem and aimed for a practical, cost-effective solution. Before we wrap up this section, Dr., can you walk me through a scenario where you used multimedia tools or AI-driven methods—like predictive modeling or data visualization—in your work to communicate findings or support a decision?

I never used any air driven methods like predictive modeling. So.

I think I don't want to answer this question.

No problem. You mentioned you have a strong background in data warehousing and mining—have you used any multimedia tools, like dashboards or interactive visualizations, to present your data analysis to stakeholders or students?

No problem, we'll move on. You've described a lot about guiding students' thought processes and evaluating their problem-solving in exams—can you give me a concrete example of how you structured student evaluation or handled exam duties in your industry training or mentorship experience?

So based on my mentorship experience. I want them to. Provide the hard work actually and I want the student to ask so many. Why, what, where, when? Kind of a question, so instead of just. She's simply sitting and. Whatever I am telling and they are grabbing it, that doesn't work.

OK, so I want them to ask the question and. So once we land the topic immediately we have to go for some address role problem. So once they solve problems then only they will identify the. Plus and minus about the particular topic. OK, so that's why you know I will structure the students first of all. So basically whenever they get the. Question Paper. From the invigilator or someone. They should not surprise that what kind of, you know, different questions. That you know they can expect. So whenever that question which the subject which I thought of the student. And they should be brace themselves and ready to take any kind of different questions. Based on the, you know, theoretical approach or academic experience or practical experience. I want them to enjoy and answer the question, so whether it is correct or not, that is a different. But how? How much? They're attempting the question without having any. You know, second thought. At the same time, I want them to study the DSA problems on daily basis. So that it will improve their, you know. Uh, improve their? Quality of thinking. And how they are handling the problems and solve it will improve their problem solving skills. Particularly. So this is how I will structure and mentor the students from my end.

Right, so you value an active, questioning mindset over passive learning. What’s one method you’ve used to actually measure whether your trainees or mentees are developing that kind of critical thinking in real time?

That's a pretty high-level approach—just to clarify, have you actually created or graded formal assessments yourself, or has your role been more of a mentor guiding students to prepare for these exams rather than directly handling exam duties?

I don't know why are you asking this kind of a weird question, so I already told. I don't have any academic experience. Then how would I, you know? Go for the exam duty.

It's a interview call. At the same time, you should not have this kind of a foolish questions correct.

Got it, that’s clear—you haven’t performed formal academic exam duties. Let’s cover one last thing: you mentioned your strong industry background, but have you had a chance to contribute to any industry projects or consultancy work that directly bridged academia and industry, like collaborating with universities or supporting student internships?

Alright, that's clear—you haven't handled formal academic exam duties. Dr., last thing: have you published any research papers in reputable journals, or contributed to conference proceedings related to your domain?

So I have published around 4 scoopers paper and I have attended 2 conference meeting in different colleges.

This is part of the academic side and apart from that, I have published lots of white papers in my industry. Companies so that it. It's going a different level based on the technology. Which we are using current key. And. That will help them to understand. How we can approach the problem in the real world testing?

Dr., can you share the title of one of your Scopus-indexed research papers and briefly describe the main contribution or finding from that work?

You've described your PhD topic in broad terms—can you give me one specific technical challenge you faced during your customer churn prediction research, and how you addressed it?

So. In customer prediction sometime we are facing some non balanced data set. And fixing the data set, it is getting much time actually. At the same time improving the accuracy, finding the right model optimization. Took some time to build the model. So this is what I have been facing as. Important obstacle while I perform the. APST.

Right, class imbalance is a classic issue. When you were optimizing your model, did you use any specific technique—like resampling, SMOTE, or cost-sensitive learning—to handle the imbalanced dataset, or did you take a different approach?

Any sampling have taken?

Alright, Dr., we're wrapping up. Is there anything you'd like to ask about the role or the team?

I took resampling method. OK. To fix the nonlinear and imbalance data set.

Alright, Dr., we're wrapping up. Is there anything you'd like to ask about the role or the team?

I just want to know what is your role and how the team will be? So do you have any any any? A solution to that? Do you have any anything about that?

Sure. My role is Dean of Faculty and Search Committee Chair—I oversee academic hiring and support faculty development. The team you'll join focuses on teaching, mentoring, student project guidance, curriculum delivery, and research, with an emphasis on high-quality academic standards. The institution is Vellore Institute of Technology, a leading private university known for its innovative programs and strong research culture. Anything else you'd like to ask?

Goodnight, my baby. Thanks. Thanks for giving a chance.