With a dedicated focus on enhancing educational experiences, my journey as an English Teacher at  in Bishkek has been defined by a commitment to curriculum innovation and effective communication. Leveraging my multilingual capabilities, I served as a Translator, facilitating clear and precise cross-language discourse for a brief period.

Our team's approach to classroom management has fostered an environment conducive to learning and growth. By championing creative thinking and problem-solving, we've cultivated a space where students are not only educated but inspired to explore the vast landscape of language and culture.

Happy International Women’s Day! 🌸
Today I want to celebrate the incredible women in my network.
Women who build companies.
Women who write books.
Women who write code.
Women who lead teams.
Women who create powerful ideas in marketing.
Women who keep learning, growing, and creating new paths in technology, business, and beyond.
Your ideas, courage, and persistence shape the future more than you may realize.
Thank you for the inspiration, the conversations, and the strength you bring to this community every day.
Wishing you energy, curiosity, and many bold projects ahead.
Happy March 8th!
#InternationalWomensDay
#WomenInTech
#WomenInBusiness
#IWD2026

undefined undefined

In Motion.Coming soon on Amazon 👇

Not another post.
Not another project.
A book.
Soon.
#IndieAuthor
 #WritingCommunity
 #AuthorJourney
 #NewBook

This edition explores how AI is gradually expanding its role in software engineering, from architectural reasoning to debugging and maintenance workflows. 
I look at current research on what AI can assist with today and where human engineering judgement remains essential.

#AI #SoftwareEngineering #ArtificialIntelligence #SystemArchitecture #TechResearch #JuliaValois #AmazonAuthor

Software engineering interviews are entering a strange transition phase.
Some companies ban AI during interviews.
Others encourage it.
But very few actually explain how engineers should use AI without looking like they’ve outsourced their brain.
This creates a new kind of skill requirement.
The engineer of today is not someone who refuses AI.
 And not someone who blindly copies from it.
The real skill is different:
• using LLMs to accelerate your work
 • understanding the code the model generates
 • questioning it
 • correcting it
 • explaining the reasoning behind the solution
In other words- you still have to be in the driver’s seat.
People who have observed thousands of technical interviews say the same thing:
 a new type of interview is emerging  one where AI is part of the process.
What we’re really witnessing is the emergence of a new type of engineer.
Not the one who writes every line alone 
 but the one who can think, reason, and lead the machine.
The interview process hasn’t fully caught up yet.
But the direction is already clear.
#AI #SoftwareEngineering #LLMs #TechCareers #FutureOfWork

Just shared a short music piece inspired by the book I’m currently writing.
If you happen to be wearing headphones right now, I’d genuinely love your take, does the sound capture the mood? 🎧
#WritingJourney #BookInProgress #SoundDesign #CreativeProcess #IndieAuthor

Software engineering is quietly changing


For decades software engineering was a deeply human craft.

Developers used logic, experience and intuition to translate requirements into code, design architectures and maintain systems over time. Writing code, debugging and designing interfaces were core human activities.

Today something more subtle is happening.

Engineering capability is gradually shifting from humans to machines.

Modern language models can already generate functions, suggest modules, explain error messages and propose APIs. Some systems can even fetch data, generate code, run tests and prepare changes for review with minimal human supervision.

This is not about machines becoming conscious.

It is about the slow transfer of practical engineering competence.

In small tasks the impact is already visible. Developers can generate functions or modules faster than before. Yet the results are mixed. In some situations work accelerates. In others it slows down because engineers spend time crafting prompts and reviewing generated output.

Another interesting shift is happening in how developers choose technologies. Tools that are easier for AI systems to reason about are becoming more attractive. Clear typing systems and structured APIs give models guardrails, making it easier to generate correct calls and interfaces.

At the same time, natural language is increasingly becoming an input for software creation. Describing an application in plain English can now produce working front-end code and integrate it into an existing development workflow.

Large-scale repository analyses suggest that a significant portion of new code is already generated by AI systems. Productivity may increase, particularly for experienced engineers exploring new domains.

But there is also a trade-off.

More code is being produced, yet less code is being carefully refactored. The long-term maintainability of rapidly generated systems is still an open question.

The real shift is therefore not automation of individual tasks.

It is the gradual relocation of engineering capability itself.

Software engineering may be moving from a craft performed entirely by humans to a hybrid process where humans guide systems that increasingly perform parts of the work.

How do you see this shift affecting the future role of software engineers?

#softwareengineering #ai #technology #futureofwork #development

AI made me 5× faster.
But it also forced me to rethink what engineering actually is.


A few months ago I started using AI in my daily workflow.

The change was immediate.

Tasks that used to take hours suddenly took minutes.
Refactoring became easier.
Debugging became faster.
Even complex merges started to move much more smoothly.

At some point I had an uncomfortable thought:

If one engineer becomes five times faster… how many engineers does a company really need?

Technology has gone through similar shifts before.

When higher-level programming languages appeared, engineers stopped writing most code at the machine level.
Automation later removed large amounts of manual operational work.
Infrastructure tools eliminated many tasks that once required entire teams to manage servers.

Each time the same pattern repeated.

Productivity increased.
The number of people needed for certain tasks decreased.
But engineering itself did not disappear.

Instead, the work moved to a higher level.

Working with AI makes this very visible.

Because the hardest parts of engineering are rarely the lines of code.

The difficult parts are things like:

• deciding what a system should do when requirements are unclear
• balancing performance, reliability, cost and security
• coordinating people who want different things from the same system
• diagnosing strange behaviour in real environments
• taking responsibility when systems fail

Those moments require judgement.

And judgement does not come from pattern completion.

It comes from experience, context and responsibility.

AI can generate excellent code.

But engineering has never been only about code.


Lesson

AI will dramatically increase productivity.

But the engineers who remain valuable will not be those who type code the fastest.

They will be the ones who can:

• navigate uncertainty
• design systems rather than isolated components
• coordinate people and decisions
• take responsibility for real-world outcomes

In many ways the role of the engineer is shifting.

Less time writing code.
More time understanding systems, trade-offs and consequences.

And interestingly, using AI every day is what made me realise that.

What part of engineering do you think AI will struggle with the most?

#AI
#Engineering
#SoftwareEngineering
#FutureOfWork
#SystemThinking
#Technology

The rollup debate is often framed as “ZK vs optimistic.”
But the real story may be about economics and architecture.
Current ecosystem data shows that optimistic rollups still dominate by secured value, while validity-based systems remain smaller economically. At the same time, the technological boundary between the two approaches is becoming less clear. Some optimistic designs are already incorporating cryptographic proof mechanisms, which gradually blurs the distinction between the categories at the infrastructure level.
Another shift comes from the economics of data availability. Recent protocol upgrades significantly increased the capacity for publishing rollup data and reduced its cost. Historically, data availability represented a large share of rollup transaction fees. As this component becomes cheaper, the cost structure changes.
A new constraint begins to emerge: proof generation.
If data availability becomes inexpensive, the performance and decentralization of proving infrastructure may become the factor that differentiates architectures.
At the same time, rollup stacks are becoming more modular. Different layers handle data availability, sequencing, interoperability, and settlement. Some interoperability approaches increasingly rely on proof verification rather than re-executing transactions across chains.
Sequencing is another active design area. Many rollups still rely on centralized sequencing, which raises questions around censorship and liveness. Current research explores alternatives such as shared sequencing layers or models where transaction ordering is performed by the validator set of the base layer.
As rollup architectures mature, the main bottlenecks appear to be shifting. Instead of focusing primarily on data availability, future designs may be shaped more by proving infrastructure, sequencing models, and interoperability mechanisms.

If data availability continues to become cheaper, which layer will become the true architectural constraint for rollups: proving, sequencing, or interoperability?
#blockchain #rollups #distributedcomputing #systemarchitecture #web3research

How to stay relevant as engineering shifts toward AI-assisted workflows 🤔 
Over the next decade, engineering workflows will change significantly.
Routine tasks that once took hours: log analysis, basic mocks, test generation, scaffolding are increasingly handled by AI systems.
This does not eliminate engineering work.
It changes where the human effort goes.
Instead of writing every integration manually, engineers are gradually moving toward supervising systems, designing architectures, and orchestrating tools that generate or execute parts of the workflow.
At the same time, the technical landscape is becoming more complex.
Modern organisations operate hundreds of services and SaaS applications.
Each of them exposes APIs, events, and data pipelines that need to interact reliably.
AI agents are also beginning to interact with systems through APIs, increasing traffic and requiring new interface layers designed for automated systems.
This creates a paradox:
while routine work is automated, the number of systems that need to be connected and governed keeps growing.
In my view, staying relevant in this environment means shifting attention from narrow implementation tasks to broader system thinking.
Engineers who remain valuable tend to focus on things that automation struggles with:
• understanding how heterogeneous systems interact
 • designing resilient integrations and observability
 • reasoning about ambiguous or undocumented behaviour
 • ensuring reliability, privacy and governance across data flows
 • structuring APIs and interfaces that both humans and automated agents can use safely
The technical stack will continue to evolve.
But the core skill that survives every transition is the ability to understand systems and make them work together.

Don’t compete with automation on routine tasks.
Move toward architecture, interoperability and system reasoning.

How are you adapting your engineering skills as AI tools become part of everyday workflows?
#softwareengineering #integration #systemsengineering
 #platformengineering #api #aiworkflow #interoperability

I still remember the smell of new floppy disks.


FORTRAN.
Pascal.
The first time I launched DOOM.
Later 
Heretic.

MS-DOS blinking on a black screen.
A stack of floppy disks with Verbatim written on them.
A Pentium I that felt like the future.

Back then computers felt mechanical and understandable.
You typed commands.
The machine responded.

Technology moved forward but slowly enough that you could feel every step.

Today I opened my email and read something that made me pause:

«“4,000-person AI replacement at Block…
70% of Web3 professionals would now rather bet their careers on small or mid-sized teams (less than 50 people) than a legacy giant.”»

For the first time, it genuinely unsettled me.

For years I thought the idea that AI would replace engineers was mostly hype.

But the latest models feel… different.
Stronger than I expected.
Stronger than me in some tasks.

And that realization did something strange in my head.

The fiction book "IT cabin" that  I’m writing  the one I treated as imagination, as speculation about the future suddenly started to feel less like fantasy and more like something that might actually happen.


From floppy disks →
to massive tech companies →
to AI systems writing code.

The arc suddenly feels very short.

Technology rarely moves the way we predict.

Sometimes the stories we write as fiction quietly turn into the first drafts of reality.


Do you still remember the first moment technology made you feel both excited and uneasy at the same time?

#AI #Technology #Web3 #Startups #FutureOfWork #TechHistory

The new kind of pressure in software development

Something subtle is changing in engineering teams.
AI tools make many tasks faster.
Code appears quickly. Prototypes appear faster than before.
And naturally, expectations start moving with that speed.
Deadlines shift.
Iterations shorten.
Teams begin to assume that more can be done in the same amount of time.
But speed created by AI is not always the same as human speed.
Generating code is fast.
Understanding systems, designing architecture, and making careful decisions still takes time.
Many engineers are quietly adjusting to this new rhythm:
working alongside tools that move instantly, while real engineering still requires thought.
It’s not necessarily a crisis.
But it is a new type of pressure that didn’t exist before.

AI accelerates production.
Engineering still requires reflection.
The challenge for teams is learning how to balance both speeds.

Have you noticed changes in expectations around speed since AI tools became part of everyday development?
#SoftwareEngineering
 #AITools
 #EngineeringCulture
 #DeveloperLife
 #FutureOfWork

taplio

Julia I.'s Linkedin Analytics

Julia I.

Performance ScoreJulia I.'s LinkedIn Performance

Julia I.'s Best Posts (last 30 days)

Ready to grow your LinkedIn presence?

taplio

.css-6ytdur{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;}Julia I.'s Linkedin Analytics

Julia I.

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);}Julia I.'s LinkedIn Performance

Julia I.'s Best Posts .css-1emoo84{font-size:var(--chakra-fontSizes-md);font-weight:400;color:var(--chakra-colors-gray-600);}(last 30 days)

Ready to grow your LinkedIn presence?

Julia I.'s Linkedin Analytics

Performance ScoreJulia I.'s LinkedIn Performance

Julia I.'s Best Posts (last 30 days)