Paid cohorts, courses are really good.

The thing is not everyone is ready to spend on them.

So I've compilted free courses(without certification) from Harvard

Python, Artificial Intelligence & Datascience courses covered
 

{ Introduction to Artificial Intelligence with Python }

 
https://www.edx.org/course/cs50s-introduction-to-artificial-intelligence-with-python

{ Introduction to Data Science with Python }

https://pll.harvard.edu/course/introduction-data-science-python?delta=0

{ CS50's Introduction to Programming with Python }

https://www.edx.org/course/cs50s-introduction-to-programming-with-python

{ Data Science: Machine Learning }
https://pll.harvard.edu/course/data-science-machine-learning

{ Data Science: Productivity Tools }

Keep your projects organized and produce reproducible reports using -- --- GitHub,
-- Unix/Linux
-- RStudio.

https://pll.harvard.edu/course/data-science-productivity-tools

{ Introduction to Probability }

https://www.edx.org/course/introduction-to-probability?webview=false&campaign=Introduction+to+Probability&source=edx&product_category=course&placement_url=https%3A%2F%2Fwww.edx.org%2Fschool%2Fharvardx

Free Harvard Python Courses

"How to find duplicate data in SQL ?"

Its a very commonly asked question in interviews, especially at junior to mid level.

Here are 5 different ways

{ Using GROUP BY & HAVING Clause }

• This method is effective for finding duplicate records based on specific columns 

• Group the data by the columns you want to check for duplicates

• Use the HAVING clause to filter groups with more than one record

{ Using Self Join }

•  A self-join can be used to compare rows within the same table
 
• Join the table with itself based on the columns you want to check for duplicates

• Filter the results to find rows where the primary key values differ but the other columns are identical

{ Using ROW_NUMBRE() }

• This method assigns a sequential number to each row within a partition

• Create a partitioned result set using ROW_NUMBER() 

• Filter rows with a ROW_NUMBER greater than 1 to identify duplicates

{ Using EXISTS Clause }

•  The EXISTS clause can be used to check for duplicate records based on certain conditions 

• Create a subquery to find potential duplicate records

• Use the EXISTS clause to check if a matching record exists in the subquery

{ Using DISTINCT }

•  While not directly for finding duplicates, DISTINCT can be used to identify unique values in a column

• Count the total number of rows

• Count the number of distinct values in the target column

• If the counts differ, duplicates exist

5 Ways to Find Duplicate Data in SQL

Service Mesh, Clearly explained !!!

Topics covered
• 3 major challenge with microservices architecture
• Service mesh with sidecar pattern
• Control plane & sidecar proxy 

3 major challenges in microservices architecture are

• How to services find other services & how to ensure newly added services are discoverable instantly ?

• How to ensure service within a cluster/team are able to authenticate each other to maintain security ?

• How to ensure monitoring of service health of so many individual microservices ?

What is sidecar pattern ?

• The sidecar pattern is a design pattern that adds supporting features to a primary application by attaching a separate container, known as a sidecar, to it. 

• The sidecar container runs in the same context as the primary application and enhances its capabilities without requiring any modifications to the service code

Service mesh with sidecar pattern

•  The service mesh is formed using a control plan & a sidecar proxy

•  The control plane is responsible to automatically injecting sidecar proxies into all the pods

What is sidecar proxy ?

The job of the sidecar proxy is to perform non business logic tasks such as •     •  Service discovery
•  Load balancing
•  Metrics
•  Security
•  Retries
•  Telemetry

It basically abstracts our all the non business logic which can be commonly written & injected into all microservices by the control plane

Developers can also focus on business logic whilst the sidecar proxy code is handled elsewhere

Ultra HQ Image available on my Github

https://github.com/jaydeepkarale/SystemDesign

I write no BS stuff about Python, System design & AI/ML.

You can stay up to date with my work on my free newsletter by subscribing to it https://jaydeepkarale.substack.com/ 

Demystifying Service Mesh: Challenges and Sidecar Patterns

If you are a beginner this is the most common SQL interview question

"How to find duplicate data ?"

Here are 5 different ways

Service Mesh clearly explained !!!

Topics covered
• 3 major challenge with microservices architecture
• Service mesh with sidecar pattern
• Control plane & sidecar proxy 

UUIDs vs Serial values for Primary Keys/row identifiers in a database ?

Image Source ~ @PlanetScale 

What is UUID ?

> Short for Universally Unique Identifier 

> 36-character alphanumeric string 

> Often used  to identify rows of data within a database table, with each row assigned a specific UUID

> Example of a UUID: acde070d-8c4c-4f0d-9d8a-162843c10333

What is Serial value ?

> A simple autoincremented sequence starting from 1 for first row, 2 for second row & so on

Problem with Serial value ?

> Consider a ecommerce database with Orders table where order_id as row identifier. Having a serial value for order_id makes it easier to guess the next value

> A REST API call /orders/<order_id> easily becomes vulnerable to attacks

Using UUID solves this problem by uniquely identifying each row with a randomly generated 36 character string make is impossible to guess the other data rows.


Using UUIDs does have it's own challenges such as

> More storage required & hence using it for storage critical apps can be an issue

> Indexing will take longer

What are your thoughts on using UUIDs as row identifiers ??? 

Choosing Between UUIDs and Serial Values for Database Primary Keys

With over half of our waking hours spent at work, the workplace isn’t just where we earn a living—it shapes our mental, physical, and emotional well-being

Choosing a good workplace with promotes healthy work life balance should be non negotiable for everyone.

Some insights based on experience that may help you in making a better decision when choosing the next opportunity 

{ Mental Health Impact }

High-stress environments can lead to burnout, anxiety, and depression

Supportive workplaces promote mental well-being through policies like counseling, mental health days, and stress management programs

{ Physical Health Considerations }

Sedentary office environments, poor ergonomics, and long working hours can negatively impact physical health, leading to issues like back pain, eye strain, and fatigue

Companies that offer wellness programs, fitness challenges, or access to gyms can improve employees’ physical health

{ Work-Life Balance }

Overworking leads to exhaustion, less time for personal life, and deteriorates relationships outside of work

A healthy work-life balance ensures employees have time to recharge, resulting in higher productivity and satisfaction

Look for workplaces offering flexible schedules, remote work options, or time-off policies that prioritize personal time

{ Supportive Leadership }

Good managers promote a balanced work environment by encouraging reasonable workloads and respecting boundaries

Leadership that listens and acts on employee feedback regarding workload and work environment can significantly improve overall well-being

{ Social Environment and Culture }

A toxic work culture with poor communication, lack of recognition, or office politics can affect mental health

Workplaces with inclusive, supportive cultures help foster collaboration and a sense of belonging

{ Workplace Autonomy } 

Employees with more control over their tasks and schedules tend to feel more empowered and less stressed

Micromanagement can lead to frustration and reduced job satisfaction, affecting mental health

{ Opportunities for Growth } 

A workplace that encourages learning, development, and career growth can boost morale and mental engagement

Lack of growth opportunities can make employees feel stuck and unmotivated, affecting overall happiness

{ Workplace Environment }

A safe, clean, and comfortable work environment contributes to better focus, productivity, and overall health

Exposure to natural light, proper ventilation, and comfortable seating are essential for physical and mental well-being

{ The Importance of Boundaries }

A company that respects personal time, including off-hours communication policies, promotes healthier work habits

Clear boundaries between work and personal life reduce the risk of burnout and improve job satisfaction

Choosing a Healthy Workplace

Python supports multiple inheritance i.e. one class can inherit from multiple classes

This leads to a problem when both parents implement same method & child get's confused which parent classes method it should call

This is where Method Resolution Order (MRO) comes into picture 

• Method Resolution Order describes a clear order in which a child class will look for methods in the hierarchy  

• MRO can be accessed using __mro__ or mro()

•  MRO helps developers write maintainable and understandable code when dealing with complex class hierarchies

If you prefer learning about MRO in video format checkout my YT video about the same

https://youtu.be/EcIMnwOTiTA?si=v5shxJNRjfyjpDV6

Understanding Python's Method Resolution Order

mTLS ensures secure communication between clients & servers !!!

Let us learn how !!!

• SSL released in mid 1990s was used to secure HTTP requests but was soon replaced with TLS

• TLS 1.3 introduced in 2018 is still the defacto standard for encrypting communications

• TLS works by sending a certificate signed by an certificate signing authority (CA)

•  The major limitation of TLS is however that it verifies the identity of the server but not the client

• If the CA is compromised, it could be used to issue fake certificates & hence impersonate the server 

•  mTLS short for Mutual TLS works by adding an extra layer of the client also authenticating against the server

• This results in more certificate management overhead but the gain in security is significant

 • mTLS finds uses cases in microservice security, regulatory compliance & preventing man-in-the-middle(MITM) attacks