Amazon

Software Development and Architecture for Ring. 
- Owning multiple critical cloud-based services that help customers protect their homes and control them from their phones with devices such as video doorbells, home security cameras or alarm systems.
- Transitioning these services into modern architectures and Amazon's ecosystem after Ring's acquisition

Software Development Engineer II

Software development and architecture for Amazon Retail.
- I designed and led the implementation of a scalable service to expand the AddToCart feature on the Search page to 90% of the Search Traffic
- I led multiple iterations of expansion of AddToCart feature in Search page, amounting to over $1bn yearly revenue
- I was part of the design review group of the large InTech organization, contributing to technical designs from any team in the org

Software development and architecture for Amazon retail

- Designed and built a new page template coordinating with 3 partner teams,
integrating in 2 services and managing our own infrastructure.
- Designed and led an iteration with small and CX changes of a 1M daily impressions layout toggle
- Implemented read use case and GDPR compliance for new service supporting high requests/s (20k avg, 45k max)
- Unblocked a project getting wrong business data for 1 month caused by a bug in network retries in another service
- Reduced with an in-memory cache the backend latency of a 1-column mobile layout for hardlines from +2ms to -13ms
- Wrote a postmortem on the importance of finding/asking for clear API specifications of APIs you consume

Software Development Engineer I

https://strategizeyourcareer.com

I'm a Software Development Engineer at Amazon. 

1:1s with my managers were not enough for me to explore all the career topics I care about. So I decided to start writing.

You can solve the customer need with just an MVP. 
The rest can come later.
#softwareengineering #coding #programming

Francisco Manuel Soto Ramírez

SDE2 @Amazon | Author of 19,000+ subs Newsletter (strategizeyourcareer.com)

No human nor AI can hold the full complexity of a system in their head.

That's why abstractions exist. They let you think in terms of “send a message” instead of sockets, TCP, retries, packet loss... 

But abstractions always leak. At some point the hidden details show up:

* Your ORM forces you to debug SQL.
* Your cloud function fails because you hit cold starts.
* Your caching layer surprises you with stale data.

Every abstraction trades speed of building for loss of control. 

That’s why great engineers don’t just use abstractions. They go one level deeper, so when the leaks appear, they aren’t drowning.

♻️ REPOST to help someone today
#softwareengineering #coding #programming

Most engineers think they have to choose between being a generalist or a specialist. Both paths sound safe, but both come with weak spots.

↳ Generalists know a little about everything but struggle to earn credibility when stakes are high
↳ Specialists go deep in one area but get stuck in a narrow lane and miss the bigger picture
↳ Generalists get overlooked, specialists get trapped

The better approach is to become T-shaped. 

Go deep in one area so people trust you when systems break, and build broad knowledge so you can connect dots across teams, tools, and business priorities. 

Think about it like a videogame. Maxing out one stat leaves you useless in most fights. Spreading all stats evenly keeps you average forever. The winning build is depth plus breadth.

Companies reward T-shaped developers. Depth gives you authority. Breadth makes you adaptable. Together, they create the leverage that drives promotions and career growth.

👉 You can become a T-shaped developer here: https://lnkd.in/dBfQVQng

♻️ REPOST to help someone today

#softwareengineering #coding #programming

Productive engineers don't waste code review time with obvious mistakes.

Run your code through AI first. It will catch the big issues, style slips, and consistency problems.

That way, engineers can spend their review time on what really matters: architecture, edge cases, and long-term maintainability.

AI handles the noise. Humans handle the judgment.

Learn more about using AI for code reviews: https://lnkd.in/dFi-TYhm

♻️ REPOST to help someone today
#softwareengineering #coding #programming

Do you think you lack technical knowledge?

For 99% of engineers, that's not the case. 

Most people think they need to be a technology specialist, but that limits your career.

Tomorrow's newsletter breaks down the T-shaped developer model and how it creates faster growth than being a generalist or specialist.

Here is what you will learn:
↳ Why pure specialists and pure generalists both stall in their careers
↳ How T-shaped engineers create influence across teams and leadership
↳ Practical ways to deepen your expertise while broadening your scope
↳ How to apply skill stacking daily at work

📣 Subscribe here to get tomorrow's insights: https://lnkd.in/gJcEpjzQ

#softwareengineering #coding #programming

.armchair {
 border: 5px solid black;
}

Tag below your frontend developer friend
#softwareengineering #coding #programming

𝐁𝐞𝐟𝐨𝐫𝐞: No engineer used AI. Those who used it had an advantage.

𝐍𝐨𝐰: Everyone uses AI. The advantage is context engineering.

I wrote an article explaining what context engineering is, and the different techniques to do it.

I share how I've cut code review time with AI in my real workflow

I'll also share:

↳ Why context engineering matters more than smarter models
↳ Techniques to make AI output faster and more accurate
↳ Practical steps to get started without lots of infrastructure and costs.

If you want to move past generic AI use and start showing measurable results, this one is for you.

👉 Join 20,000+ engineers reading the full article here: https://lnkd.in/dFi-TYhm

♻️ REPOST to help someone today

#softwareengineering #coding #programming

Most engineers ruin their Monday on Friday.  

They close the laptop with 10 open loops.  
No reflection. No plan.  

Result? Monday morning starts with chaos.  

Instead, before logging off, do this:  
- Write down what's not finished.  
- Reflect on what you accomplished
- Set 3 priorities for Monday.  

That way, Monday isn’t ambiguous.  
You'll open the laptop and get to work.  

☝️ I created an Engineer's Workstation to become the most productive engineer on Earth: https://lnkd.in/dQhnPzJB

#softwareengineering #coding #programming

The way we build software hasn't really changed since 1960... until now.

Most engineers are using AI to write code, but the best ones are using it for more than that. 

In tomorrow's newsletter, I talk about the two new ways to build software that are changing everything.

Knowing when to use each is the new skill for career growth. 

📣 Subscribe to get the full breakdown: https://lnkd.in/gJcEpjzQ

#softwareengineering #coding #programming

Complaining is much easier than doing it yourself.

But I’ve found the best software engineers are the ones who complain through actions.

↳Missing docs → They write docs
↳Missing tests → They always write tests
↳Missing metrics → They add metrics

Which kind of software engineer do you choose to be?

♻️ REPOST to help someone today
👉 Follow Francisco Manuel Soto Ramírez for more
#softwareengineering #coding #programming

If you're a software engineer, you have to read these 54 books:

↳ Cracking the Coding Interview
↳ Designing Data-Intensive Applications
↳ Clean Code
↳ The Pragmatic Programmer
↳ Fundamentals of Software Architecture: An Engineering Approach
↳ Building Microservices
↳ Grokking Algorithms
↳ Design Patterns: Elements of Reusable Object-Oriented Software
↳ Software Engineering at Google
↳ The Effective Engineer

The rest are here: https://lnkd.in/dYaf4J9C

#softwareengineering #coding #programming

All software engineers should move faster than is comfortable

If you ship once a month, you can hide sloppy architecture behind QA cycles

If you ship once a quarter, no one notices if your deployment ir rolled back a few times until it succeeds...

But if you ship five times a day, you are forced to become a real software engineer

↳ Your tests need to actually protect you
↳ Your observability must show seconds after something breaks
↳ Your deploys must be atomic and reversible
↳ Your architecture must handle restarts without downtime

Move fast, and the speed itself will police your engineering quality

Do you agree or disagree?

♻️ REPOST to help someone today
👉 Follow Francisco Manuel Soto Ramírez for more
#softwareengineering #coding #programming

If your career was an RPG, how would you spend your skill points?

Most engineers make one of two mistakes:

→ Put all points into one stat → Deep specialist, unstoppable in one fight, useless everywhere else.
→ Spread points evenly across all stats → Jack of all trades, but never unlocks advanced abilities.

Both paths cap your growth.

The winning strategy? Build a T-shaped character.

→ Max out one core skill (backend systems, ML, frontend performance).
→ Invest the rest in supporting stats (system design, communication, product sense, leadership).

That stack unlocks hidden powers:

↳ Authority. People trust you as the go-to expert.
↳ Adaptability. You connect dots across teams.
↳ Influence. You drive org-wide progress, not just local code.

This is exactly how promotions work in FAANG.

It's not about typing code faster.

It’s about solving hard problems, aligning teams, and creating business impact.

You can read a full article about the T-Shaped developer here: https://lnkd.in/dBfQVQng

#softwareengineering #coding #programming

Most engineers burn out by giving 100% when things are calm.

Then a big launch or incident hits, and they have nothing left.

The fix:
↳ Leave breathing room (don’t plan every hour)
↳ Plan energy, not hours (like capacity planning)
↳ Use calm times to recharge: take trainings, explore doing automations...

Productive engineers aren’t the ones at 100% every day.
They’re the ones ready to give 100% when it matters.

♻️ REPOST to help someone today
👉 Follow me for more
#softwareengineering #coding #programming

Most people are tired of AI messing up their code. They only knew vibe-coding (aka prompt engineering), and they couldn't make their AI behave.

But there's a solution, a new paradigm. It's called Spec-Driven Development:

🔹 𝐏𝐫𝐨𝐦𝐩𝐭 𝐄𝐧𝐠𝐢𝐧𝐞𝐞𝐫𝐢𝐧𝐠 is fast and great for small, uncertain tasks as long as you know how to ask sharp questions
🔹 𝐒𝐩𝐞𝐜-𝐃𝐫𝐢𝐯𝐞𝐧 𝐃𝐞𝐯𝐞𝐥𝐨𝐩𝐦𝐞𝐧𝐭 starts by asking AI to write specs, stories, and designs before you touch implementation

Fixing a test? Keep it casual with prompts.

Building a feature that others will depend on? Start with a spec.

👉 Join 19,800+ engineers reading the full article here: https://lnkd.in/dxexJG-N

♻️ REPOST to help someone today
#softwareengineering #coding #programming

I was closing the most tickets, but it was quietly making me feel stagnant.

It feels productive, right? Clearing the board, shipping features. 

You're busy, but not necessarily making progress on what matters for the next level. The difference between a good engineer and a top-tier one isn't just coding skill. It's the ability to operate with intention.

Here's a simple habit you can start tomorrow to break the cycle. 

↳ Define a few goals for your next quarter
↳ Define 3 daily activities that feed into those goals.
↳ Repeat every single day

This small habit is part of a larger system called multi-scale planning. 

👉 This full article explains the framework and contains a template to start implementing it immediately: https://lnkd.in/dQhnPzJB

♻️ REPOST to help someone today

#softwareengineering #coding #programming

Prompt engineering vs spec engineering in less than 1 minute:

Prompt engineering 

↳ Conversational, iterative coding with AI
↳ Best for small, scoped tasks (bug fixes, refactors, tests)
↳ Uses inline prompts like “add null check” or “optimize query”
↳ Prioritizes speed, but decisions live in chat history
↳ Fragile on big projects, easy to lose context

Spec engineering

↳ Start by prompting AI for a plan (requirements, design, tasks)
↳ Review the spec before code generation
↳ Best for features, services, or long-lived systems
↳ Provides auditability and team alignment
↳ Slower upfront, but reduces rework and hallucinations later

The "senior" skill isn’t choosing one. It’s knowing when to switch.

♻️ REPOST to help someone today
#softwareengineering #coding #programming

Most engineers start each AI tool with a blank page. That’s wasted effort.

Productive engineers carry their context across apps. They don’t rewrite the same goals, notes, or prompts 10 times.

Now standards like https://agents.md/ make it possible without changing formats.

The real productivity boost isn’t the AI model. It’s how well you engineer your context.

☝️P.S. That's called context engineering, and I wrote this article about context-engineering: https://lnkd.in/dFi-TYhm
#softwareengineering #coding #programming

LinkedIn

li_at

JSESSIONID

Vibe-coding got a lot of hype because it feels fast and exciting, but it doesn’t scale.

We shouldn't try to replace the software lifecycle with AI, but improve each stage with it.

Spec-Driven Development beats vibe-coding every time because it applies AI across:

↳ requirements
↳ design
↳ coding
↳ code review

This means better software without burning out your team chasing vibes

PS. I went deeper in this article if you’re into this kind of thing: https://lnkd.in/dxexJG-N

Most teams are using AI to skip steps:

They end up destroying the systems that made them great in the first place
They let juniors vibe-code entire services from scratch
They ask ChatGPT to architect without constraints

The smartest teams use AI to sharpen what already works, not replace it

↳ Code review becomes tighter and faster using automated suggestions
↳ Spec generation turns loose ideas into precise engineering plans
↳ Test writing goes from boring chore to near instant
↳ Bug hunting becomes proactive and AI assisted

Keep your development lifecycle intact

Use AI inside a Spec-driven process

☝️ I wrote a full article about vibe coding vs Spec-driven development here: https://lnkd.in/dxexJG-N

♻️ REPOST to help someone today
#softwareengineering #coding #programming

Taplio

Francisco Manuel Soto Ramírez's Linkedin Analytics

What is Francisco Manuel talking about?

's Best Posts (last 30 days)

Want to drive more opportunities from LinkedIn?

Taplio

.css-1jgreyo{background-image:linear-gradient(284deg, rgba(39,43,72,1) 0%, rgba(33,47,183,1) 100%);color:transparent;-webkit-background-clip:text;background-clip:text;}Francisco Manuel Soto Ramírez's Linkedin Analytics

What is Francisco Manuel talking about?

's Best Posts .css-qqfgvy{font-size:var(--chakra-fontSizes-md);}(last 30 days)

Want to drive more opportunities from LinkedIn?

Francisco Manuel Soto Ramírez's Linkedin Analytics

's Best Posts (last 30 days)