SDE & AI @Amazon | Simplifying software engineering, system design, and generative AI | Building Hungry Minds: the newsletter to stay ahead in software 👇

Tell me about the last time you felt growing as a software engineer. Really.

𝙽𝚞𝚕𝚕𝙿𝚘𝚒𝚗𝚝𝚎𝚛𝙴𝚡𝚌𝚎𝚙𝚝𝚒𝚘𝚗? Here's why:

We all face information overload as software engineers. 
There's too much content and too little time.
You have to stay relevant, but it's hard.

Right now, you're either working or taking a break.

You are hoping to get a dopamine shot from scrolling your feed.

You can find:

💡 A practical tip to implement
📚 An inspirational quote
😆 Something funny

And then what?

→ Scroll past it? 
→ Forward it blindly to a friend?
→ Forget about it in the bookmarks?

The harsh truth: nothing.

What you consume passively doesn't make you learn.

How do I know?
It was the same for me for years.

How do you stay updated with AI, stay relevant in the job market, and continuously grow?

That's the problem I set out to solve with Hungry Minds - a free weekly newsletter to help software engineers learn and stay updated, without the noise.

I'm doing that by:

1. Writing to you daily on LinkedIn to 40K+ followers and over 5M impressions last year
2. Compressing 1 week of software deep dives and trends in 1 free newsletter issue
3. Sharing my thoughts and learnings being promoted to SDE II in 1.5y at Amazon

If that sounds interesting, check out Hungry Minds or DM me! I'm always down for a chat.

Save 50+ hours of your week here: 𝗵𝘁𝘁𝗽𝘀://𝗵𝘂𝗻𝗴𝗿𝘆𝗺𝗶𝗻𝗱𝘀.𝗱𝗲𝘃 (6K+ engineers from startups to big tech are reading already)

DISCLAIMER: Opinions expressed are solely my own and do not express the views or opinions of my employer.

My career background:

I'm currently an L5 software engineer at Amazon in Madrid, working on core Search CX and Personalization. 
In 2022/2023, I delivered 2 successful billion-scale projects with multiple teams across different time zones.

I have worked with multiple code languages (although this doesn't make a difference), and with tech stack in the lines of AWS CDK, React applications, scheduling systems, backends with GraphQL, and simple AI applications.

My education background:

- I finished a Master's in Machine Learning and AI with l'Ecole Polytechnique and IPP
- I completed a Master's in Computer Science and Data Science with ESILV Paris
- I got a Master's in business development and marketing with EMLV Paris

Alexandre Zajac

𝗖𝗼𝗱𝗲 𝗱𝗼𝗰𝘂𝗺𝗲𝗻𝘁𝗮𝘁𝗶𝗼𝗻 𝗶𝘀 𝗡𝗢𝗧 𝗔𝗯𝗼𝘂𝘁:


→ One senior gate-keeping all the knowledge. 

Code documentation is ALL about:

→ Sharing knowledge so the whole team knows how the system works.

1-person docs = single point of failure.

Whole-team docs = redundancy + knowledge multiplication.

Remember:
You NEED documentation to manage a system through people and time.


How do you manage docs effectively?


P.S: Documente my personal interview cheatsheet: https://lnkd.in/d_eCGEJR

SDE & AI @Amazon | Building Hungry Minds to 1M+ | Daily Posts on Software Engineering, System Design, and AI ⚡

No alternative text description for this image

10+1 System Design Patterns Every Engineer Should Know in 2026:


___
btw, my personal interview cheatsheet: https://lnkd.in/d_eCGEJR
___


0.💡Load Balancer: 

Distributes incoming traffic across multiple servers to ensure no single server gets overwhelmed.

➡️ Useful when: You're scaling web applications or ensuring high availability.

1.💡CDN (Content Delivery Network):

Caches content at edge locations worldwide, reducing latency by serving users from the nearest node.

➡️ Useful when: You're serving static assets globally or streaming media at scale.

2.💡Message Queue:

Decouples services by letting them communicate asynchronously through a queue (like RabbitMQ or Kafka).

➡️ Useful when: You need reliable communication between microservices or handling traffic spikes.

3.💡Rate Limiter:

Controls the number of requests a user or service can make within a time window.

➡️ Useful when: You're protecting APIs from abuse or managing resource allocation.

4.💡Circuit Breaker:

Prevents cascading failures by stopping requests to a failing service and allowing it time to recover.

➡️ Useful when: You're building resilient microservices that depend on external APIs.

5.💡Database Sharding:

Splits a database into smaller pieces (shards) distributed across multiple servers.

➡️ Useful when: Your database is too large for a single server or you need horizontal scaling.

6.💡Caching Layer:

Stores frequently accessed data in memory (Redis, Memcached) to reduce database load.

➡️ Useful when: You're optimizing read-heavy applications or reducing latency.

7.💡API Gateway:

A single entry point that handles routing, authentication, and rate limiting for all your microservices.

➡️ Useful when: You're managing multiple services and need centralized control.

8.💡Reverse Proxy:

Sits in front of servers and forwards client requests, adding security, caching, and load balancing.

➡️ Useful when: You need SSL termination, compression, or hiding server architecture.

9.💡Event-Driven Architecture:

Services react to events rather than direct calls, enabling loose coupling and real-time processing.

➡️ Useful when: Building real-time analytics, notifications, or audit logging systems.

10.💡CQRS:

Separates read and write operations into different models, optimizing each for its specific purpose.

➡️ Useful when: You have complex domains with different read/write performance needs.


What did I miss?

Solving system design problems is HARD.

1 mega-template to solve any problem:


Approaching system design problems can be harder than coding problems.

For one reason: ambiguity.

13 concepts to use when solving a system design problem:


0. Load Balancer:

A load balancer efficiently distributes incoming network traffic across multiple servers to ensure no single server bears too much load.

1. API Gateway:

An API Gateway acts as a single entry point for all client requests, providing a unified interface to various microservices.

2. Cache:

Caching stores recently accessed data for speedy retrieval, reducing the load on databases and improving response times.

3. CDN:

A Content Delivery Network (CDN) stores static content closer to users to minimize latency and speed up content delivery.

4. Distributed File Storage:

Distributed file storage ensures data is redundantly stored across multiple locations for fault tolerance and easy access.

5. Search Index:

Using a search index like Elasticsearch allows for scalable and fast search capabilities across large datasets.

6. Notification Service:

A notification service handles the delivery of messages or alerts to users across different platforms and devices.

7. Data Processing Systems:

Data processing systems like Hadoop and Spark analyze large volumes of data, enabling scalable and complex computational operations.

8. Distributed Scheduler: 

A distributed scheduler orchestrates and manages the execution of distributed tasks across various computing resources.

9. Metadata Server:

The metadata server stores information about the structure of data, enabling efficient access and management.

10. Block Server:

Block servers manage the storage blocks in distributed file systems, which helps in data replication and retrieval.

11. Distributed Logging:

Distributed logging collects logs from various services in a centralized manner for monitoring and troubleshooting.

12. Replication and Partitioning:

Replication ensures data availability and durability, while partitioning helps in scaling databases by distributing data across different nodes.


Reference (aff.):

🔥 Grokking the system design interview: 
https://lnkd.in/dcMM8sZM


What did I miss?


~~~

👉🏻 Join 50,001+ software engineers getting curated system design deep dives, trends, and tools (it's free):
➔ https://lnkd.in/dkJiiBnf

~~~

If you found this valuable:

👨🏼‍💻 Follow Alexandre Zajac
🔖 Bookmark this post for later
♻️ Repost to help someone in your network

11 Algorithms Every Engineer Should Still Know in 2026:


0.💡 Breadth-First Search (BFS): 

Explore a graph level by level, starting from the root, which is great for finding the shortest path in unweighted graphs. 

➡️ Useful when: You're designing web crawlers or analyzing social networks.


1.💡 Two Heaps: 

Uses a min-heap and max-heap to manage dynamic datasets efficiently, maintaining median and priority. 

➡️ Useful when: You need to manage a priority queue or dynamic datasets.


2.💡 Two Pointers: 

This technique takes 2 points in a sequence and performs logic based on the problem.

➡️ Useful when: You are implementing sorting or searching functions.


3.💡 Sliding Window: 

Optimizes the computation by reusing the state from the previous subset of data. 

➡️ Useful when: You're handling network congestion or data compression.


4.💡 Depth-First Search (DFS): 

Explores each path to the end, ideal for situations that involve exploring all options like in puzzles. 

➡️ Useful when: You're working with graph structures or need to generate permutations.


5.💡 Topological Sort: 

Helps in scheduling tasks based on their dependencies. 

➡️ Useful when: You are determining execution order in project management or compiling algorithms.


6.💡 Merge Intervals: 

Optimizes overlapping intervals to minimize the number of intervals. 

➡️ Useful when: Scheduling resources or managing calendars.


7.💡 Backtracking: 

It explores all potential solutions systematically and is perfect for solving puzzles and optimization problems. 

➡️ Useful when: Solving complex logical puzzles or optimizing resource allocations.


8.💡 Trie (Prefix Tree): 

A tree-like structure that manages dynamic sets of strings efficiently, often used for searching. 

➡️ Useful when: Implementing spell-checkers or autocomplete systems.


9.💡 Flood Fill: 

It fills a contiguous area for features like the 'paint bucket' tool. 

➡️ Useful when: Working in graphics editors or game development.


10.💡 Segment Tree: 

Efficiently manages intervals or segments and is useful for storing information about intervals and querying over them. 

➡️ Useful when: Dealing with database range queries or statistical calculations.


What did I miss?

You are what you eat.

10 brain foods to grow as an engineer:


0. Why Clock Synchronization Makes Engineers Lose Sleep:
► https://lnkd.in/gkDbWARS

1. How ChatGPT Actually Stays on Track (by Neo):
► https://lnkd.in/ggNMHDuV

2. How to Build Internal AI Agents That Actually Work:
► https://lnkd.in/gkJKyQnm

3. Scaling LLMs to Larger Codebases:
► https://lnkd.in/gWwfibJQ

4. Connection Pooling Finally Explained in 6 Minutes (by Nikki):
► https://lnkd.in/gkx6eNSG

5. 5 Engineering Dogmas It's Time to Retire (by Anton):
► https://lnkd.in/gYTdTmap

6. The System That Eliminates Interruptions for 10x Engineers (by Francisco):
► https://lnkd.in/g65_i6jf

7. Why Logging Sucks (And Everyone Knows It):
► https://loggingsucks.com/

8. Internal Dev Tooling at Meta (from Gergely):
► https://lnkd.in/g_A_BxFd

9. The Future of Software Development Is Software Developers:
► https://lnkd.in/gvzpK2CM


I am packaging the other deep dives and trends in today's issue of Hungry Minds.

Join 50,001+ software engineers from big tech to startups here: https://lnkd.in/dkJiiBnf


What are you reading this week?


#softwareengineering #systemdesign #programming

CTO | Author of Engineering Leadership newsletter (65k+ subscribers) - Helping you become a great engineering leader!

I am an engineering leader with a passion for solving complicated problems with innovative concepts. I love learning about new technologies, sharing my knowledge and teaching others. 

I have a proven record of successfully managing Agile teams, setting the correct direction and culture as well as creating right processes and continuously improving them. I have overseen hiring and building up the teams as well as have strong emphasis on coaching and mentoring to direct reports.

You can reach me at info@gregorojstersek.com.

Gregor Ojstersek

In today’s AI era, being a great engineer is no longer defined by your individual contribution alone.

Being a great engineer is defined by the ability to make others around you better, to deliver projects successfully, and to create impact that moves the whole organization forward.

There are many ways to multiply your efforts:
- making your whole team better,
- to work well across different teams,
- be a multiplier through systems,
- multiply and scale through people.

These are all archetypes of multipliers, where each provides immense impact in its own specific way.

Learn more in this Engineering Leadership article: https://lnkd.in/eDyTMzes

CTO | Founder of Engineering Leadership newsletter (180k+ subscribers) - Helping you become a great engineering leader!

This is what happens when you do 10k lines of code per day with AI :)

Happy Friday and have a good weekend!

The speed of writing code has never been an issue.

You can write 10,000 or 20,000 lines of code, but if you are building the wrong thing for the user, it means nothing.

Once you know you’re building the right things, then you can bring in AI tools to speed up the delivery.

This is far more important than just typing faster or producing more code.

While most teams are bragging about how fast they ship code, your focus should be somewhere else:

Focus on building the right things, first.

Then, use technology to accelerate the rate at which you produce new value.

That's how you win in tech in 2026.

What makes a great engineer in the AI-era?

As tools for development got better, the desired skills for engineers have moved from just pure knowing programming languages and frameworks (expected) to problem-solving abilities and people skills.

It's less about the tasks, but how big an impact you can create.

And engineers who can create the biggest impact have the multiplier mindset.

It's the mindset where you focus less on just your pure individual tasks, but you focus on how you can make things better for your whole team, organization, and, of course, the customers (users).

Learn more in this Engineering Leadership article: https://lnkd.in/ezWydu8T

Pragmatism is one of the most important concepts in Software Engineering.

There are endless variables and trade-offs we need to consider when building Software. And a lot of them are people-focused.

And pragmatism is just going to be more and more important as time goes on. 

The reason?

AI can just predict the next best token, but what is missing is the touch of real-world experience and the feel for what may be the best decision in a certain moment, based on many different signals.

To learn how to become pragmatic as an engineer, read this Engineering Leadership article: https://lnkd.in/eGh5s5Tm

“My work speaks for itself”

Assuming that everyone knows that you are doing good work is in most cases, going to be a wrong assumption.

There are 2 things I like to say.

1. You can do the best work in the world, but if the expectations are not managed correctly, it will not be perceived that way.

2. Doing great work is only half of the work, making sure that everyone knows about it is the other half. You need to do both.

Learn how to build a successful engineer <> manager relationship in the latest Engineering Leadership article that we did together with Shehab Abdel-Salam here: https://lnkd.in/dhJ4qAyD

So simple, yet overlooked.

What you can do is limited. But a team’s impact scales.

Real impact comes from being a multiplier.

Learn more in this Engineering Leadership article: https://lnkd.in/ezWydu8T

As an engineer, I didn't properly use 1:1 meetings with my manager.

I didn't use my 1:1 meetings for my career growth, instead, I used them to give a status update.

I was missing out on so many different learning opportunities, which I could use the 1:1 meeting for:
- I could talk about career goals,
- work through plans for the promotion or
- ask for specific coaching and feedback situations.

Lesson learned: 
It’s a lot better to use 1:1 meetings to help you be successful in your role and with your overall career growth.

P.S. To learn more -> join engineers and leads from Meta, Google, IBM, Revolut in my cohort-based course here: https://lnkd.in/exsYfM54 (24 hours left to sign up!).

Meetings don’t create alignment. Decisions do.

If no one leaves the room knowing:
- what was decided
- who owns it
- what happens next

It wasn’t a meeting, it was a calendar placeholder.

Make sure to respect people’s time.

End meetings with decisions, or don’t have them at all.

Learn how to create meetings that engineers will love in this Engineering Leadership article: https://lnkd.in/ekZcdgGu

Pramoda Vyasarao went from being laid off at Meta to doing an MBA in Solopreneurship.

Meta laid off 11,000 people in January 2023, and his position as Head of Data Engineering was one of them.

He took a 2-month break to reflect on next steps, and that's when he got an idea and asked himself: What if I take a two-year break and do something on my own?

This is what he did:

1. His first product: one-on-one coaching
This was the easiest place to begin for him. He had done this at Oracle and Meta as an organizational leader. He didn’t need preparation or certification.

2. Cohort courses
Within two months, he realized one-on-one coaching doesn’t scale. That’s when he thought of online cohort courses. He started with one course and then added 3 more.

3. Digital courses
To scale further, he converted all four courses into digital products. To save time, he hired contractors to record his live sessions and professionally edit them.

4. Published a book
He published a book called Beyond Your Limits with the endorsement of Dr. Marshall Goldsmith.

5. Started to write a newsletter
He started posting on LinkedIn in March 2023. Later, in a consulting session with Justin Welsh, he convinced him to start a weekly newsletter.

After 2 years, he returned to tech, but not the same person.

Returning to corporate life felt different for him. He no longer defines himself by a title. He carries the mindset of an entrepreneur, comfortable with ambiguity, focused on impact, and willing to take calculated risks.

To learn about his full story, read the latest Engineering Leadership article: https://lnkd.in/ds6TWcvD

Coinbase scaled their hiring to 150 engineers per month over the pandemic.

How did they do it?

1. They standardized the searches

Previously, for every new hire, it would be required to draft a “job spec”. With standardization, you could only hire a frontend engineer, a backend engineer, a crypto engineer, an infra engineer, or an ML engineer (and a few others).

2. They moved from hiring in SF only to nationwide

And later to 10+ countries. This resulted in a massive increase in talent pool size.

3. Interviewing became centralized

Candidates weren't screened by their future hiring manager or peers anymore, but rather by any Coinbase engineer. This massively opened up the interviewer capacity.

4. From hiring manager round-ups to centralized hiring panels

Each day with a rotation, one Coinbase leader would host a hiring panel (with 5-6 senior managers), where all the interview packets feedbacks were reviewed and decided which ones would move to offer and which ones would be rejected.

This gave them the ability to process 40-50 candidates at the pre-offer stage a day.

5. The offers became standardized and non-negotiable

No back and forth with the candidate. Both base salary and equity were directly related to the level you would be hired in, no exceptions allowed.

To learn more, make sure to not miss tomorrow's Engineering Leadership newsletter.

Luca Bonmassar, Former VP of Engineering at Coinbase, is sharing a lot of interesting insights on how they scaled hiring.

To not miss tomorrow's article, make sure to join 180k+ engineering leaders here: https://lnkd.in/eNSKBMUf

Trust is earned and not given. 

You build trust by consistently doing the right things.

And that’s especially important to focus on when you get hired by a new company or you come to a new team. 

DON’T believe that a title can give you trust by default.

The first thing you need to do is to earn trust from others. And it takes time to do that.

But the funny thing? 

It takes a long time to build trust and it takes a couple of seconds to lose it. 

To learn how to build trust, read this Engineering Leadership article: https://lnkd.in/exFUZ4Uz

Structure stories with CARL instead of STAR for a behavioral interview.

This is what Austen McDonald recommends, and here is his reasoning.

CARL stands for:
Context: The background of what was happening and why it mattered
Actions: The specific steps you took (plural, not singular)
Results: What changed and the impact you made
Learnings: What you learned or would do differently

The key difference is that CARL combines Situation and Task into Context, eliminating the mental energy wasted trying to separate them.

And it adds Learnings, which gives you a platform to demonstrate the depth of wisdom that separates senior engineers from less experienced ones.

Learn all about how to nail behavioral interviews in the latest Engineering Leadership article: https://lnkd.in/eN5qJy7g

taplio

The Battle

Alexandre Zajac

Gregor Ojstersek

Followers

Posts

Engagements

Likes

Performance ScoreAlexandre Zajac's LinkedIn Performance

Performance ScoreGregor Ojstersek's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

Alexandre Zajac

Gregor Ojstersek

Followers

Posts

Engagements

Likes

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Alexandre Zajac's LinkedIn Performance

Performance ScoreGregor Ojstersek's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScoreAlexandre Zajac's LinkedIn Performance