Welcome to Home Page
Welcome to our blog, where technology, culture, and education converge! Join us as we explore how these dynamic forces shape our world and spark meaningful discussions.

Monday, 14 April 2025

The Transformative Power of Generative AI in Medical Chatbots

The Transformative Power of Generative AI in Medical Chatbots

Generative AI, specifically large language models (LLMs), are currently changing healthcare chatbots in ways that allow for catching up the limitations of traditional cartable based systems. LLMs excel at understanding complex language, something necessary in healthcare, since patients' questions are often vague and ambiguous. Unlike rigid rule-based chatbots, LLMs - trained on very large datasets - understand how to interprete human language, even when patients speak in non-medical, colloquial terms. Comparatively, using this improved understanding enables chatbots to respond more effectively and meaningfully without the responses being pre-scripted and is used, medical, information of the given inquiry. LLMs can review and analyze all of the large amounts of medical knowledge, as well as, the vast amounts of patient interaction data used to enable LLMs to respond in a specific manner for uniquely wanting to yield the most tailored response for a patient's needs, patients responded queries, medical history, symptoms and preferences, an experience in mental health using LLMs can identify the emotional tone in relation to the patient's questions to develop a more empathetic interaction to establish trust and raise patient engagement. All of these processes of resolving to the generative AI could allow patients to have more effective communication in their therapy session; this ability reduces the learning gap between medical terminology/knowledge and layman's terms knowledge provides patients with clearer and understandable information, allowing patients to become more informed and develop better communication skills with their health practitioners.


Enhancing Contextual Awareness and Clinical Utility


One significant improvement that Gen AI brings to the table is its awareness of context within a conversation. Most chatbots forget information and conversations become discontinuous. LLM's have an ability to remember context which creates a natural and continuous conversation. In healthcare, this seamlessness in conversation is extremely important, since it typically involves intricate conversations with much detail. Even better, Gen AI enhances the clinical usefulness of health care chatbots because it not only helps provide context but can provide even initial evaluations of patients. By training LLMs in medical guidelines and protocols, chatbots can assist in patient triaging, suggest possible diagnoses, and determine next steps. This function helps provide more streamlined healthcare delivery, increases efficiency, allows for more timely interventions, and improved patient outcomes. The ability to provide initial views can help significantly lessen some of the burden on health care providers so that they are free to focus on more complex cases.


Informing RAG Intent Chatbots for Medical Applications


To effectively provide a Retrieval-Augmented Generation (RAG) intent chatbot for medical purposes, two aspects need to be developed well: clear intent categories and a sound knowledge base. Intent categories are developed through defining user needs, categorizing them in a hierarchical order, and providing a range of training data for each intent. For example, intents could be "find information on a condition", "schedule an appointment", or "get medication advice". Afterwards, for each intent, we use several sample queries to ensure the chatbot understands the request from the user. The knowledge base which is critical for giving useful responses, involves identifying sources which include medical databases, documentation, and APIs. This information must be organised so that it is easy to use and accurate, complete, and searchable and indexed using complex search algorithms.


AI-Powered Enhancements for Medical Chatbot Functionality


AI, particularly in Natural Language Processing (NLP), greatly enhances medical chatbot capabilities. Chatbots can precisely detect patient intent, recognizing the nuances of patient questions and not just key words. AI is also responsible for retrieving appropriate medical data fromas of medical data and literature around the world. It allows patients to receive factual and relevant, timely data so that they may stay up to date and aware of the consequences of medical decisions due to their conditions. In addition to providing accurate and timely retrieved data from all data sources, by processing context-sensitive and patient-specific information, AI creates a personalized feedback experience, such as medical history, medication and allergies. Furthermore, AI trains chatbots with verified medical knowledge, ensuring medical accuracy, and teaches chatbots with new medical research on a continuous basis. In addition to these accuracy upgrades, AI provides context-awareness letting chatbots retain context during conversations creating natural and intuitive interactions. These enhancements increase the capacity of medical chatbots as helpful resources for healthcare personnel and patients, improving access to medical information and services, efficiency, and patient outcomes.


To understand human language is the essence of the chatbot. This project applies basic natural language processing (NLP) methods such as tokenization and Term Frequency-Inverse Document Frequency (TF-IDF) vectorization. With these processes, it is possible to take the text data and produce useful numerical representations that can be used as input features for the Logistic Regression model. The classification model is trained to classify the user inputs into one of the selected intents so that the chatbot can respond intelligently depending on the conversation. This project represents the first step in developing a more sophisticated conversational agent. The project has only imagined naive models with a less complex dataset, but it has established an initial foundation for future work - which may include more complex and larger datasets, more advanced and deeper machine learning algorithms, and more advanced NLP methods (ex. Named Entity Recognition (NER), sentiment analysis, context-sensitive dialog management). Overall, the major learning goals of the project are:


  • Understand how chatbots process natural language inputs to identify user intents and extract relevant information.

  • Apply text preprocessing techniques, including tokenization and TF-IDF vectorization, to prepare textual data for machine learning models.

  • Train and evaluate a Logistic Regression model to classify user inputs into specific intents.

  • Design and deploy a responsive and interactive chatbot interface using the Streamlit web framework.


View the project in my GitHub

Real-Time Interaction: The Foundation of Chatbot Utility

Real-time interaction functionality is at the center of all chatbot functionality. This inherent property enables users to engage, query, resolve problems, and seek information in real-time without the lags present in other support mechanisms. Employing natural language processing (NLP) and artificial intelligence (AI) chatbots can understand what the user is saying and respond appropriately and swiftly, and even engage in conversations on topics other than business. The immediate availability of support enhances the overall surfing experience and reduces user frustration, particularly when using services during off-business hours. The capacity to offer instant help is a stepping stone to user satisfaction and one of the major drivers of chatbot adoption in different industries.


History Storing: Enhancing Personalization and Efficiency

Another key feature that adds significant value to chatbot systems is the feature to retain previous conversations. By retaining prior interactions, chatbots can offer a more applicable and effective support experience. For users, it means they do not have to repeat the same issues or provide similar information multiple times, so the whole interaction is more efficient. For companies, having past conversations is also valuable, and the information it contains is essential to understanding the usual customer issues and allows organizations to identify possible areas for improvement and to better customize support. Finally, retaining past interaction can make it easier to switch over to a human agent, when necessary, with continuity and consistency to resolve problems.





Contact Tab: Bridging the Gap Between Automation and Human Assistance

Chatbots are effective at handling simple queries and giving automated help, but sometimes there are circumstances that require a human agent. In order to accommodate this, the contact tab should be present. It is the easily visible and easily accessible means by which end users can contact a human support representative through a variety of channels such as email, telephone or live chat.  An embedded contact tab in the chatbot interface allows companies to make sure users can easily hand off tasks that are not straightforward or complex while providing an easy flow from automated support to human support. The blended approach uses the structured and often systematic support of chatbots and combined with human agents and their depth of thinking and approaching problems, the user experience is improved in a user-focussed approach.


Thank you!

Friday, 28 March 2025

Becoming a Microsoft Student Ambassador presents the possibility to be part of technology, innovation, and leadership among students. Being an energetic person who would love to impact change, I feel this is a role where I would be able to harness my talents, grow personally, and assist in developing technology within my university and society.

Microsoft is a brand that has always fascinated me because of its vision to change the world using technology. As a Student Ambassador, I would be more than happy to act as the link between Microsoft's tools and the students I am surrounded by. Whether it is presenting students with the new innovations in software, promoting Microsoft 365 usage, or organizing cloud computing and coding workshops, I wish to encourage my colleagues to use Microsoft's powerful tools to their fullest potential.

Also, I have a proactive attitude and great communication skills that are useful for this position. My experience in planning campus events, heading student clubs, and working on projects will enable me to work efficiently in advocating for Microsoft's cause. I look forward to being part of a community of like-minded people who are passionate about technology and problem-solving. Through networking with professionals and gaining access to exclusive resources, I hope to create an environment of innovation and development among my peers.

Finally, I am excited to be able to represent Microsoft in the best possible manner and enable students to reach their full potential using the latest technology. I look forward to the opportunity to be able to assist Microsoft's mission of enabling people and organizations to do more.

Monday, 17 March 2025

Git and Github

Git and GitHub are basic tools for modern software development, providing a means of implementing version control and collaboration and facilitating the whole development process. It's a distributed version control system that lets developers see and manage the changes that happened in the codebase over time. It tracks all changes made to the codebase, allowing developers to roll back previous versions, work more efficiently, and record their project history. This works locally on a developer's computer, allowing the person to work separately and synchronize their changes with a central repository afterward. 



On the other hand, GitHub is a cloud service hosting Git repositories, where developers can collaborate on their work more easily and share code with others. It offers a social layer over Git where developers can create public or private repositories, manage issues, process pull requests, and work with other people on open-source projects. Besides this, it also provides wikis, project boards, and continuous integration features to extend development.The key strength of Git revolves around working with branches where developers may handle individual features or fixes before merging them into the master codebase. This allows for minimizing code conflicts while allowing parallel development. 

Alongside this, GitHub allows easy pull request handling, through which developers can propose changes into a project and negotiate them with the team members. In short, the union of Git and GitHub provides a considerable version control and collaboration arsenal. Git enables fine-grained control over changes, while GitHub offers a platform for sharing and collaboration with other developers on code. Together, they constitute the very backbone of modern software development.


Learn more


Saturday, 8 March 2025

Create machine learning models

Machine learning (ML) is rapidly advancing technology that makes recommendation engines or driverless cars possible. Developing machine learning models is one complex procedure for deriving solid predictions from raw information. This involved several important steps along the way. The first step is data collection. It is important that the data is relevant and of good quality, as it is going to train any ML model. This data can be harvested from databases, sensors, or APIs. Outliers or irrelevant data should be cleaned and preprocessed. This may include normalization, encoding categorical variables, and splitting the data into training and testing sets.

 After the data has been properly prepared, the next phase is selecting an appropriate machine learning algorithm. The selected algorithm is dependent on the type of problem—that is, classification, regression, clustering, or recommendation. Some of the most commonly used algorithms include decision trees, linear regression, support vector machines (SVM), and neural networks. For example, decision trees are often used in classification problems, while linear regression is what you'd find in predicting a continuous value. The next step is model training, which incorporates training data after an algorithm has been selected. 

During training, the model learns the patterns hidden in the data. Also, performance metrics, such as the accuracy, precision, recall, or mean squared error, can be used in assessing the performance of the model. Hyperparameters are those parameters that regulate the behavior of the algorithm, and if tuned well using techniques that include cross-validation, the best performance of the model will be achieved. Finally, once models are properly trained and optimized, they get deployed for real-world applications for prediction on new data. The model continues to be monitored and maintained to make sure it learns as conditions in the data change over time. In conclusion, the process of machine learning model development is a structured one, involving proper data handling, algorithm identification, building the model, and deploying the model into the real-time prediction application. 

Link for more information


Introduction to Visual Studio Code for Education

Visual Studio Code is a popular open-source code editor developed by Microsoft that has excellent flexibility is easy to learn and use, and offers numerous features. While it has inarguably taken the lead among professional developers, it has also been given considerable recognition in the education space by revolutionizing the learning and teaching processes of programming. In that case, the free VS Code presents an opportunity through accessibility, simplicity, and universality of high-cost customizing its features, which is an extremely valuable choice for learning programming in different kinds of institutions of learning. 

One of the key sweeter benefits of using VS Code in education is the lightweight nature of the application. Unlike the more hefty IDEs, VS Code installs quickly, runs fast across a plethora of devices, and supports many programming languages such as Python, JavaScript, C++, and Java. It enables students to write, run, and debug code in a single environment, allowing for easy learning thanks to features such as integrated terminals, source control integration, and debugging capabilities. Another intriguing feature of VS Code is the large catalog of extensions that allows students to change and manipulate their coding experience with version control tools, such as linters, code formatters, and even living collaboration tools associated with those efforts. 

Extensions to GitHub and Docker make it possible for students to work together on real projects, mimicking professionals' tools and practices. In addition, completion rules and highlighting syntax of the text editor allow better practice of bug-free code and efficient learning of programmers. Moreover, VS Code works cross-platform and thus runs on Windows, macOS, and Linux, ensuring easy use by different student populations. Moreover, the open-source nature of it enables teachers to design environments for particular learning objectives for a personalized and inclusive learning experience.

 Link for more information



Introduction to Teams meetings and calling

Microsoft Teams has now cemented its position among the essential communication and collaboration tools of today. With remote and hybrid workspaces becoming the most sought-after ways to work, organizations are trying to garner effective and stable communication platforms. Teams calling and meetings keep people and organizations connected and provide seamless collaboration anytime, anywhere. Teams meetings, as the core of the platform, provide a very simple yet efficient way of conducting virtual meetings. Microsoft Teams allows meetings, which can range from small-scale team check-ins to large-scale corporate conferences, through video, audio, and screen sharing, adding to real-time collaboration.


 An added advantage during the meeting is the possibility for real-time collaborative document sharing and editing that is inherently knitted in with Microsoft 365 applications such as Word, Excel, and PowerPoint. Teams meetings league an entire bouquet of experience features: breakout rooms, live captions, and recording. Besides meetings, Teams also comes standard with end-to-end calling capabilities like making a voice and video call directly in the Teams platform. Extra features like call forwarding, voicemail, and call transfer really turn Teams into a complete communications solution that's ready to replace legacy PBX phone systems for businesses. 

As Outlook is its sibling in Microsoft Office, Teams can schedule calls from a calendar event or an email for the communicators to ease the bulk of communication. Incentive after security, Teams also focuses on compliant communication by means of proud features, for example, end-to-end encryption and uptight supervision, like HIPAA and GDPR compliance norms. All these features make it worth joining hands for organizations looking for a safe and effective means to communicate. In summary, calling and Teams meetings are the fundamental functions that guarantee easily digestible communication and cooperation. Being the greatest with an intuitive interface, sophisticated features, and rugged security, Microsoft Teams is the subject leading-edge platform for online communication solutions. 

Link for more information


Monday, 17 February 2025

The Potential of Generative AI to Solve Global Challenges

 Since generative AI can create new content, it holds immense potential to assist in solving most of the world's biggest challenges.1 Applications in many fields are diverse and offer promising solutions to intricate issues.

In the field of medicine, generative AI will accelerate drug discovery by creating new molecules with desired characteristics that can lead to therapeutics or cures for diseases.3 Personalized medicine is advantaged through aggregation and integration of patients' clinical information, enabling the development of a customized treatment plan.4 Consider AI developing personalized vaccines or therapies based on a person's own genetics.

Climate change is another field where generative AI can be a game changer.6 It will assist in inventories of the interventions one could propose to avoid wider climate change effects by virtue of simulating climatic processes to observe the effect of said interventions.7 AI can further propose new materials that render renewable energy technologies efficient and hence inexpensive.

Food sovereignty is a global issue, and generative AI weaponizes food security by optimizing crop yields and creating sustainable agriculture practices.9 AI can scrutinize the soil quality, weather factors, and crop health to give each businessman personalized advice on how to boost productivity and reduce wastage of resources.

Generative AI can also redefine education.11 It would create personalized learning for students, where varying content and pacing are aligned with their own requirements.12 The AI tutor would offer tailored feedback and guidance, which would render education more accessible, fair, and effective.

The generative AI can serve as a bridge to bridge the communication gap, facilitating real-time translation that is community-friendly as well as the creation of accessible content for people with disabilities.14 It can also boost creativity and facilitate innovation by offering tools for artists, designers, and even scientists to hypothesize an alternative path to tackle the limits of their disciplines.

Though there are still challenges, such as those of ethical utilization and countering bias, the potential for generative AI to be applied to even broader open global issues is enormously powerful.16 If we can do so responsibly, we are committing to a more sustainable, fair, and prosperous future for everyone.


Check my LinkedIn Account

LinkedIn