Technology: 2025

Wednesday, October 22, 2025

Powershell script to bring news from news Feeds and create a nice looking HTML page.

# ===== DailyNews.ps1 =====

# RSS feed URLs

$techFeeds = @(

"https://techcrunch.com/feed/",

"https://feeds.arstechnica.com/arstechnica/index"

)

$propertyFeeds = @(

"https://www.abc.net.au/news/feed/52278/rss.xml"

)

# Output HTML file

$outputFile = "C:\temp\DailyNews.html"

# Current date

$date = (Get-Date).ToString("dddd, MMM dd yyyy")

# Function to fetch feed items

function Get-FeedItems {

param($url)

Write-Host $url

try {

$rss = [xml](Invoke-WebRequest -Uri $url -UseBasicParsing -ErrorAction Stop).Content

$items = @()

foreach ($item in $rss.rss.channel.item | Select-Object -First 5) {

$items += [PSCustomObject]@{

Title = $item.title

Link = $item.link

PubDate = $item.pubDate

Source = $rss.rss.channel.title

}

}

return $items

} catch {

Write-Host "⚠️ Could not fetch: $url"

return @([PSCustomObject]@{

Title = "⚠️ Could not fetch"

Link = $url

PubDate = ""

Source = ""

})

}

}

# Fetch items

$techNews = @()

foreach ($feed in $techFeeds) {

$techNews += Get-FeedItems $feed

}

$propertyNews = @()

foreach ($feed in $propertyFeeds) {

$propertyNews += Get-FeedItems $feed

}

# Build HTML content

$html = @"

<html>

<head>

<title>📰 Daily Tech & Property News - $date</title>

<style>

body { font-family: Segoe UI, Arial; background: #f9fafb; color: #333; margin: 20px; }

h1, h2 { color: #2a7ae2; }

a { text-decoration: none; color: #0078d7; }

a:hover { text-decoration: underline; }

section { margin-bottom: 40px; }

.article { margin: 8px 0; }

footer { margin-top: 40px; font-size: 12px; color: #777; }

</style>

</head>

<body>

<h1>📰 Daily Tech & Property News</h1>

<p><b>Date:</b> $date</p>

<section>

<h2>🏠 Property News</h2>

<ul>

"@

foreach ($item in $propertyNews) {

$html += "<li class='article'><a href='$($item.Link)' target='_blank'>$($item.Title)</a></li>`n"

}

$html += @"

</ul>

</section>

<section>

<h2>💻 Tech News</h2>

<ul>

"@

foreach ($item in $techNews) {

$html += "<li class='article'><a href='$($item.Link)' target='_blank'>$($item.Title)</a></li>`n"

}

$html += @"

</ul>

</section>

<footer>Generated automatically by PowerShell – $date</footer>

</body></html>

"@

# Save HTML

$html | Out-File -FilePath $outputFile -Encoding utf8

Write-Host "✅ HTML News Summary saved to $outputFile"

Tuesday, September 2, 2025

Springboot - Authenticating Microsoft Token

A JWT has 3 parts (Base64URL encoded, separated by dots):


<header>.<payload>.<signature>



Create a POSTMan POST request to get MS token:

POST https://login.microsoftonline.com/<tenant-id>/oauth2/v2.0/token

Send the below params in x-www.-form-urlencoded:


client_id: <id of application>
client_secret: <generated secret of that app>
score: api://<tenant-id>/.default
grant_type : client_credentials


The response would be :

{
    "token_type": "Bearer",
    "expires_in": 3599,
    "ext_expires_in": 3599,
    "access_token": "eyJ0eXAiOiJKV1QiL................."
}



Copy that access_token and get ready for actual API request:

https://<applicationurl>/<apiname>/<operationname>

For ex
https://tap/tcaservice/myapi

Send that access token as Bearer Token under Authentication Header

Here is the code for Token  validation

@Component
public class MicrosoftTokenFilter extends OncePerRequestFilter {

    private final JwtDecoder jwtDecoder;

    public MicrosoftTokenFilter(@Value("${azure.ad.tenant-id}") String tenantId) {
        //String issuerUri = "https://login.microsoftonline.com/" + tenantId + "/v2.0";
        String issuerUri = "https://sts.windows.net/" + tenantId + "/";
        this.jwtDecoder = JwtDecoders.fromIssuerLocation(issuerUri);
    }

 @Override
    protected void doFilterInternal(HttpServletRequest request, HttpServletResponse response, FilterChain filterChain)
            throws ServletException, IOException {

        logger.info("Starting MicrosoftTokenFilter");
        String authHeader = request.getHeader("Authorization");
        if (authHeader != null && authHeader.startsWith("Bearer ")) {
            String token = authHeader.substring(7);

            logger.info(" token = " + token);
            try {
                Jwt jwt = jwtDecoder.decode(token);

                SetupAuthentication(jwt);

            } catch (Exception e) {
                response.sendError(HttpServletResponse.SC_UNAUTHORIZED, "Invalid Microsoft token");
                return;
            }
        }

        filterChain.doFilter(request, response);
    }

    private void SetupAuthentication(Jwt jwt) {

            Map<String, Object> claims = jwt.getClaims();

            String name = "";
            String email = "";
            String orgCode = "";
            boolean isAuthUser = false;
            String environment = "TEST";
            String orgName = "";

            List<Role> roles = new ArrayList<Role>();
            for ( Map.Entry<String,Object> claim : claims.entrySet()) {

                if (claim.getKey().equals("EMAIL")) {
                    email = claim.getValue().toString();
                }
            }

            //set faked
            name = "Naas api";
            email = "vkumar@austroads.gov.au";
            orgCode = "Naas";
            isAuthUser = true;
            environment = "TEST";
            orgName = "Austroads";

            if (isAuthUser) {
                User user = new User();
                user.setFirstName("");
                user.setLastName(name);
                user.setEmail(email);
                user.setOrganisationCode(orgCode);
                user.setRoles(roles);
                user.setOrganisationName(orgName);


                Authentication authentication = createAuthentication(jwt, user, environment);
                setAuthentication(authentication);


                logger.info("Authorities set in context:" + SecurityContextHolder.getContext().getAuthentication().getAuthorities());

            }
    }

    private Authentication createAuthentication(Jwt jwt, User user, String environment) {       
        List authList = new ArrayList();   
        authList.add(new SimpleGrantedAuthority("TEST_MyAdminRole");
        JwtAuthenticationToken usertoken = new JwtAuthenticationToken(jwt, authList);
        return usertoken;       
    }

    private void setAuthentication(Authentication authentication) {
        SecurityContext context = SecurityContextHolder.createEmptyContext();
        context.setAuthentication(authentication);
        SecurityContextHolder.setContext(context);
    }

}

Wednesday, August 27, 2025

Azure API Management tool

API: Application Programming Interface.

The API Gateway pattern is a design pattern used in distributed systems, especially in microservices architecture, to provide a single-entry point for all client requests. Instead of clients calling each microservice directly, they send requests to the API Gateway, which then routes, transforms, secures, and aggregates responses from backend services.

Popular tools:

Azure API Management (APIM) – what you’re learning now.
AWS API Gateway
Kong, Apigee, NGINX, Tyk, Traefik, Spring Cloud Gateway

Analogyy:

Think of an API Gateway as a reception desk: visitors (clients) arrive at one desk, state their needs, and the receptionist handles the internal coordination with different departments (other APIS) before giving the visitor a consolidated response.

Core Functions of an API Gateway:

Request routing – Directing incoming calls to appropriate backend services
Authentication and authorization – Verifying identities and access rights before allowing requests
Traffic management – Controlling request flow to prevent system overload
Response caching – Storing frequently requested data to improve performance
Request/response transformation – Modifying data formats to ensure compatibility
Logging and monitoring – Tracking usage patterns and system health

Azure API Management:

APIM is a tool by Microsoft Azure which works based on API Gateway pattern.

Core Features of APIM:

API Gateway: Acts as an entry point for API requests, handling routing, load balancing, and caching.
Developer Portal: Provides a customizable interface for developers to discover, learn about, and consume your APIs.
Management Plane: Allows API providers to manage APIs, set policies, define security measures, and monitor performance.

Components of APIM:

Tuesday, April 22, 2025

Add Emojis ti your code - Shortcut

In visual studio 2022, you can add emojis like checkmark, smile and others in the code or logs to spruce up your logs or comments.

Shortcut Windows + ;

Search and chose name .

For example 0 fire, checkmark

Monday, April 7, 2025

Authentication in .NET

Type	Common Use Case	Token Type / Auth Flow
✅ JWT Bearer Token	APIs, SPAs, Mobile Apps	Token-based (Stateless)
✅ Cookie Authentication	Web Apps (MVC / Razor Pages)	Cookie-based (Stateful)
✅ OAuth2 / OpenID Connect	Social Login, Enterprise SSO	External Identity Providers
✅ API Key Auth	Simple Public APIs	Key in Header/Query
✅ Windows Authentication	Intranet apps, Enterprise networks (AD/LDAP)	Windows/Kerberos
✅ Certificate Authentication	High-security APIs (B2B)	X.509 client certs
✅ Custom Token / HMAC	Legacy systems, Highly controlled environments	Custom logic

What does JWT Has

JWT - JSON Web Token has 3 parts

Header - metdata information

Payload - User centric inform

Signature -

Tuesday, April 1, 2025

Count colored cells in Google sheets - create custom formula

In the google sheets, use the following script and enjoy the function to count the colored cells

=countColoredCells(B59:AF59, "#274e13") #Green color code

function countColoredCells(countRange,colorRef) {

  var count = 0;

  var activeRange = SpreadsheetApp.getActiveRange();

  var activeSheet = activeRange.getSheet();
  var formula = activeRange.getFormula();
 
  var rangeA1Notation = formula.match(/\((.*)\,/).pop();
  var range = activeSheet.getRange(rangeA1Notation);
  var bg = range.getBackgrounds();
  var values = range.getValues();
    
  for(var i=0;i<bg.length;i++)
    for(var j=0;j<bg[0].length;j++)
    {
      if( bg[i][j] == colorRef )
        count=count+1;
    }   
  return count;
};

Tuesday, March 25, 2025

Creating a Dental Application Front End using Angular

1. First install Nodejs from nodejs.org

2. Download and Install VSCode

3. After installing nodejs, set path environment

4. Set the Path to nodejs

5. Then in terminal the following error occurs if we try to check "npm install" as npm is not added.

npm : File C:\Program Files\nodejs\npm.ps1 cannot be loaded because running scripts is disabled on this system. For more information, see about_Execution_Policies at https:/go.microsoft.com/fwlink/?LinkID=135170.

At line:1 char:1

+ npm

+ ~~~

+ CategoryInfo : SecurityError: (:) [], PSSecurityException

+ FullyQualifiedErrorId : UnauthorizedAccess

6. Run the following command:

Set-ExecutionPolicy RemoteSigned

7. Install Angular CLI using

npm install -g @angular/cli

8. ng new mydental

9. cd mydenal

10. ng serve mydental

11. For better UI, lets add Angular Material

ng add @angular/material

12. Create your first component

ng generate component dashboard

Saturday, January 11, 2025

AI Learning Roadmap

Learning to create AI Agents is a fascinating journey that combines elements of programming, machine learning, and problem-solving. Here's a breakdown of how you can embark on this learning process:

1. Build a Strong Foundation:

Programming Fundamentals:

Python: This is the most common language for AI and machine learning due to its readability, extensive libraries, and active community. Learn basic syntax, data structures, control flow, and object-oriented programming concepts.
Other Languages (Optional): While Python is primary, knowledge of languages like JavaScript (for web-based agents) or C++ (for performance-critical applications) can be beneficial later.

Mathematics:

Linear Algebra: Vectors, matrices, matrix operations are essential for understanding many machine learning algorithms.
Calculus: Derivatives and gradients are crucial for understanding optimization techniques used in training models.
Probability and Statistics: Understand probability distributions, hypothesis testing, and statistical inference.

2. Dive into Machine Learning (ML):

Core Concepts:

Supervised Learning: Regression, classification, various algorithms like linear regression, logistic regression, support vector machines (SVMs), decision trees, random forests, and neural networks.
Unsupervised Learning: Clustering (k-means, hierarchical clustering), dimensionality reduction (PCA), anomaly detection.
Reinforcement Learning (RL): Learning through trial and error, concepts like Q-learning, SARSA, policy gradients, and deep reinforcement learning.

Libraries and Frameworks:

Scikit-learn: A great library for general-purpose machine learning, offering implementations of many standard algorithms.
TensorFlow and PyTorch: Powerful frameworks for building and training deep learning models. Choose one and learn it well.
Keras: A high-level API that simplifies building neural networks, often used with TensorFlow or PyTorch.

Start with Simple Projects:

Handwritten digit recognition (MNIST dataset): A classic introductory project using neural networks.
Spam detection: A practical application of text classification.
Regression problems: Predicting house prices or other numerical values.

3. Understand AI Agent Concepts:

What is an AI Agent?

An agent that perceives its environment, takes actions, and learns to achieve a goal.
Differentiated from simple AI models by their ability to interact with an environment.

Key Components:

Perception: How the agent receives information about the environment (sensors, data).
Decision-Making: How the agent chooses actions (algorithms, policies).
Action: How the agent affects the environment.
Learning: How the agent improves over time based on experience.

Types of Agents:

Reflex Agents: Simple agents that react to current percepts without considering history.
Model-Based Agents: Maintain an internal model of the world.
Goal-Based Agents: Act to achieve a specific goal.
Utility-Based Agents: Maximize a utility function that quantifies the desirability of states.

Environments:

Fully Observable vs. Partially Observable: Can the agent perceive all aspects of the environment?
Deterministic vs. Stochastic: Are actions predictable?
Discrete vs. Continuous: Are the state and action spaces discrete or continuous?

4. Learn Reinforcement Learning (RL) in Detail:

The core of AI agents: RL is essential for enabling agents to learn to make decisions in dynamic environments.
Key RL Concepts:

Markov Decision Processes (MDPs): A mathematical framework for modeling decision-making problems.
Exploration vs. Exploitation: Finding a balance between trying new things and exploiting known good strategies.
Rewards and Punishments: The signals that guide the agent's learning.
Value Functions: Estimating the long-term reward of being in a particular state.
Policy: The agent's strategy for selecting actions.

RL Algorithms:

Q-Learning, SARSA: Table-based RL algorithms for discrete state and action spaces.
Deep Q-Networks (DQNs): Combining Q-learning with deep neural networks for high-dimensional state spaces.
Policy Gradient Methods (e.g., REINFORCE, PPO, A2C): Directly learning the policy.

RL Environments:

Gym (OpenAI): A toolkit for developing and comparing reinforcement learning algorithms, offering a variety of environments.
Custom Environments: Create environments tailored to your specific needs.

5. Practice and Build Projects:

Start with Simple Agents:

Grid-world navigation: An agent learning to navigate a simple grid environment.
Game-playing agents (e.g., Tic-Tac-Toe, simple games in Gym): Apply RL to train agents to play games.

Gradually Increase Complexity:

More complex game environments (e.g., Atari games using DQN).
Developing agents for more realistic tasks.
Experiment with different RL algorithms and architectures.

Contribute to Open-Source Projects: This is a great way to learn from others and get real-world experience.

6. Stay Updated:

Follow Research Publications: Keep up with the latest advancements in AI and RL.
Attend Conferences and Workshops: Learn from experts and network with others.
Online Courses and Tutorials: Resources like Coursera, edX, Udacity, and YouTube are invaluable for continuous learning.
Online Communities and Forums: Engage with other learners and experts to ask questions and share knowledge.

Key Resources:

Online Courses:

Coursera's "Machine Learning" by Andrew Ng.
Udacity's "Deep Reinforcement Learning Nanodegree".
edX's "AI Professional Certificate".

Books:

"Reinforcement Learning: An Introduction" by Sutton and Barto (free online version).
"Deep Learning" by Goodfellow, Bengio, and Courville (free online version).

Libraries and Frameworks:

Scikit-learn
TensorFlow
PyTorch
Keras
OpenAI Gym

Websites:

OpenAI
DeepMind
Google AI

Tips for Success:

Be Patient and Persistent: Learning AI takes time and effort. Don't get discouraged by initial challenges.
Focus on Understanding the Fundamentals: Don't just memorize code; strive to understand the underlying concepts.
Start Small and Build Up: Begin with simple projects and gradually increase the complexity.
Practice Regularly: The more you code and experiment, the better you'll become.
Collaborate with Others: Learn from your peers and share your knowledge.

In summary, learning to create AI agents is a multifaceted process that requires a solid foundation in programming, math, and machine learning, with a strong focus on reinforcement learning. By following this roadmap, practicing consistently, and staying updated with the latest advancements, you can embark on a rewarding journey in this exciting field!