[BDD 2025 - Mobile Development] Mobile Engineer and Software Engineer: Are we still relevant? (Sidiq Permana)

Is Software Engineering
still Relevant in
The Age of AI?
Sidiq Permana
Chief Innovation Officer of NBS.DEV
Google Developer Expert

Things to Talk
● Glance of NBS
● Problems in Software Development and Software Engineering
● How do we Software Engineering Practice in General
● LLM and AI in Software Engineering
● Are we still relevant?

NBS.DEV
● The most reliable software engineering consultant
in Indonesia.
● Consists with 136+ People and WE ARE STILL
HIRING.
● Trusted by Top Notch Industries and Technology
Leaders.
● Been delivering 100+ successful solutions and still
counting.
● Experience in “Fireﬁghter” in Software Delivery.

We create
high-quality, reliable, scalable, and sustainable
products that drive long-term business success.
Beyond get things done!

Top-notch Industry and Technology Leaders Trust to Our Services
We’ve worked with

We’ve worked with
and many more
Top-notch Industry and Technology Leaders Trust to Our Services

We are still HIRING!
CHECK THE IG
@nbs.corner

Challenge in the
Software Development
and Engineering

Software Engineering is the investment of
systematic application of engineering principles
and multidisciplinary approaches to the entire
software development lifecycle.
By leveraging modern technology, it provides a
structured framework to solve complex problems
and deliver high-quality, robust, and scalable
solutions.
At NBS we believe that

It requires a lot of soft
skills as we deal with
people and hard skill in
how we create, design
and deliver the
outcomes

How it’s
valued
Diverse Principles,
Multidisciplinary
and Best Practices
— that becomes a
structured
framework for
solving problems
Cultivated
Knowledge and
Proven Approach
that shapes from
years of hands-on
practice, challenges,
and opportunities
Blending deep,
specialized
knowledge with
honed skills, all
reﬁned through
hundred thousands
of hours of practical
experience
01 02 03
Knowledge Experience Expertise

“The software industry faces a critical challenge: a
widespread lack of attention to quality assurance,
resulting in inconsistent quality across all aspects
of the development lifecycle.

It’s a Common Terms, it should be easy to do

How AI-in the Loop
Advancement can
Address this issue?

Software and Approach
are Changing

It mutates the form
And Delivery Practice
● Software 1.0: Requires less resources, hard to build, but easy to operate
● Software 2.0: Requires mid resources, harder to build, but moderate to operate
● Software 3.0: Requires high resources, hardest to build but easiest to operate

AI in Software Engineering is able to
deliver the better solutions and
products. It’s beyond Vibe Coding

EVEN, it has well-known limitations
It can deliver a wrong
answer that even
humans can rarely
do and it doesn’t
have any soft skill
that we have.
Jagged
Intelligence
It requires a long and
bigger working
memory and
resources to nail one
context to the next
one.
Anterograde
Amnesia
It can be harmful if
we can not control it
well and causing
serious matters in
Risk, Security and
Privacy.
Gullibility
It can produce
hallucinations based
on prompt you make,
unclear prompt,
insufficient context
and lack of speciﬁc
knowledge leads to
the unexpected and
overactive results.
Hallucinations
LLM AI

Depth and
Broader
Leveraging AI for research
provides a holistic
understanding of user needs
and market dynamics,
accelerating the discovery of
opportunities previously hidden
by complexity.
The Power of LLM AI
In Software Engineering
Harnessing
Research
Well Deﬁned
and Planed
The deep, comprehensive
understanding of a complex
problem is essential for
structuring a clear solution
framework, ultimately leading
to the highest degree of clarity
in the pre-development process.
Pre-Development
More
Productive
Direct and indirect assistance
throughout the development
process, accelerating
engineering work by
automating tedious tasks and
elevating code quality through
a continuous safety net against
defects.
On-Development
By integrating into the
end-to-end DevSecOps pipeline,
It enables continuous risk
management, ensuring
product reliability and
high-quality service by
minimizing potential tradeoffs.
Post-Development
More
Resilient

Android
Lint
CI
Code Repository
Automation
Language &
Framework Backing Service
Knowledge Sharing
IDE
Android Vitals
Core Library
Mobile Engineering Development
Pipelines

Android
Lint
CI
Code Repository
Automation
Language &
Framework Backing Service
Knowledge Sharing
IDE
Android Vitals
Core Library
Mobile Engineering Development
Pipelines
V
i
b
e
C
o
d
i
n
g
V
i
b
e
C
o
d
i
n
g
Test Automation
Generation Risk and Issues
Insights
Code Quality and
Safety
Build
Automation
Deploy
Automation
Deep Research

AI Model +
Our Specialized Knowledge +
Our Code =
https://www.slideshare.net/slideshow/bdd-2025-mobile-development-mobile-engineer-and-software-engineer-are-we-still-relevant-sidiq-permana/Internal Software Engineering
Tooling that Automate Many
Things, and again, it’s beyond
Vibe Coding.

But,,,
Vibe Coding cause serious effect
for the Long Term Scalability.

Vibe Coding as AI Assisted Coding is shifts the way we
do software engineering process, the utilization of A can
become Software Engineer Partners, that brings the level
of Software Engineers to be an orchestrator during the
solution development.
Keep in mind that this shifts requires a solid of Basic
Principles and foundations that they are still remain the
same, and keep existed like algorithm, data structure,
programming principles and best practices as well as
tools, frameworks and languages.
To make the AI Utilization keeps in Control
How we see the opportunity
and Tradeoffs

Veriﬁcation
& Validation
A
I
Generating

How we can do a better
veriﬁcation if we are still
lacking of knowledge and
experience?

How we can do
better?
In AI Assisted Coding
Keep in Tight Hash

● Describe a single, concrete, incremental change that brings clarity of context and purpose
● Don’t ask for code, ask for approaches: Pick an approach, draft a code, Review and learn, Small Chunk, Repeat
● Verify and Test
● Commit Often
● Ask for a suggestion on what would be implemented next
● Repeat

IF your prompt vague then the generation
will be problematic/unexpected result then
the veriﬁcation is failed and then it starts
spinning

How can we do a better veriﬁcation if
we don’t have any sufficient knowledge
of foundation, basic principles,
understanding of the tools and
frameworks that we us?
Better Result?
Better Understanding.

Reverse
Engineering
Forward
Engineering
Analyze Systems
1. AI Based requirements,
experience design, and
rapid prototyping
2. AI based reverse engineer
Write Systems
1. Coding Agent
2. Testing Agent
3. Arch-Agent
4. Deploy Agent
5. Ops Agent

We do side by side with
AI in Software Creation
through UI/UX
Augmentation
Deliver us the result,
suggestions and
augment our actions
Agent
It connect to many
agents talk to each other
and automate some
process under our
supervision
Future of Software and Apps We Build
How it goes?
Partial
Autonomy

What if.. Building
software is no longer
bottleneck?

When AI
accelerates from
Opportunity
Identiﬁcation,
Idea to Impact

AIFSD :
AI First Software
Delivery

Engineering
Practices still
Matter

The Future of Software Engineer
1. AI takes us to the higher level as software engineering is raising the ﬂoor
and AI is raising the ceiling. An Orchestrator. The expert Generalist.
2. Soon or later software engineer becomes the full stack engineer that is
equipped with end to end understanding of the software building. A
Leadership Role.
3. It can say a typical T shape that has depth technical knowledge and
breadth the surrounding knowledge like PM, Product Development,
DevOps, Data, QA and etc. IT REQUIRES ADAPTABILITY.
4. Even you need to understand how to retrain, ﬁne tune, and do RAG or
create your own Agent. IT REQUIRES THE DESIRE TO LEARN MORE.

“Not who fast but it will belong to
who has a clear vision, think
deeply, adapt quickly, customer
focused, curious and collaborate
effectively.”

Code Responsibly
Supervise Mindfully

AI in Software Delivery : “If we do good job of categorizing
what those things are in the development life cycle, and we
can automate checks and guardrails, perhaps we can have
autonomous agents handling those functions. But with very
few exceptions, we’ll need humans is the loop. It’s not
optional.”
Martin Fowler

Takeaways
1. Software engineering is beyond the writing of code, it’s far complex
2. It has a big challenge of the lack of attention and inconsistent quality
3. Well utilization of AI in the software development and engineering
process can improve productivity and increase the bar of the High Quality
outcome
4. Tools changed but core principles and core practices last
5. Keep the ﬂame of curiosity lights up that make us, as software engineers
are able to catch up and stay relevant, at the end we are the orchestrator
and expert generalist.
6. Software Engineering in the Era of AI requires a visionary software
engineer that connects the foundational principles, tools and social skills
to build remarkable products. We Lead the AI.

Thank You
sidiqpermana10 nouvrizky10
sidiqpermana
Get in touch
sidiq@nbs.co.id

[BDD 2025 - Mobile Development] Mobile Engineer and Software Engineer: Are we still relevant? (Sidiq Permana)

More Related Content

Similar to [BDD 2025 - Mobile Development] Mobile Engineer and Software Engineer: Are we still relevant? (Sidiq Permana)

Recently uploaded

[BDD 2025 - Mobile Development] Mobile Engineer and Software Engineer: Are we still relevant? (Sidiq Permana)