At MetaCTO, we build custom mobile apps that help businesses grow new revenue streams, create online communities, increasing customer LTV with digital products, and make client communication more personal.

Christopher Fitkin

I’m genuinely worried about junior engineers.

Okay, let’s get this out of the way (I’m annoyed that I even feel the need to say this): Yes… AI is useful. (We use it every day. It makes our teams faster. That’s not the debate.)

The problem is what it’s removing from the learning process - especially for devs coming up after c. 2023.

For most of my career, junior engineers (myself included) learned by doing the hardest, most boring part of the job. The only question you were ever asked on an interview: “Can you write code.”

The grind of banging out massive amounts of code everyday is where my instincts developed. I learned what a bad abstraction felt like because I had to live with it. I learned why certain shortcuts were dangerous because I broke things.
 Then, I had to know how to fix them.

But that critical implementation stage is disappearing fast.

Today, AI can generate decent code in minutes. Which is great… unless you never learned how to reason about that code without help.

That’s the part I’m worried about.

So here’s my advice to anyone early in their career (and yes, I know this sounds old-man-y… I don’t care):

Build something on your own time WITHOUT AI. Pick a project: auth, database, API, frontend, deployment. Whatever.

You will struggle. You will make bad decisions. You’ll stay up all-night scanning thousands of lines of code until your eyes almost pop out of your head.

And that’s the point.

Because later, when you do use AI (and you should), you’ll know when it’s wrong. You’ll know when something feels off. You’ll be able to diagnose a problem & understand why a solution works.

My prediction: the juniors who put in reps (without the AI crutch) will move up the ladder faster & be the seniors all the vibe-coders rely on.

I really hope more people talk about this!

Senior dev: “Yeah, we f*cked that one up - fix it by Wed.”
New dev: “Okay so just to give some context here, this issue is actually the result of a few interconnected factors, one of which dates back to a decision that was made before I joined the team. There were also some edge cases that weren’t documented, and the requirements changed slightly after implementation. I want to be clear that the code technically works as designed, but the design itself may not have accounted for all downstream effects. I’m still investigating root cause and will circle back with findings. Open to thoughts.”

After 20+ years working in companies big and small… the fastest way up the ladder is knowing how to communicate.

They ask a question - you give an answer.
They need something done - you give a date.

No one likes/reads 7 paragraph Slack messages. 🙂

New AI security report out & it’s BAD news for Claude, Codex, Cursor, Replit, etc (or at least, cause for major concern)…

Researchers at Tenzai tested the most popular AI coding tools by having them generate common apps like shops and forums.

Result:
Every. Single. One. shipped with real security flaws.

And this wasn’t just theoretical… they discovered hackers are ALREADY exploiting:

- Permission bugs exposing other users’ transaction data
- Checkout flows that let users enter negative quantities and get paid
- Business logic that allow unlimited refunds, discounts, or credits (bc edge cases weren’t properly defined)

AI writes code fast → yes
AI doesn’t always understand incentives, abuse, or edge cases → also yes

My advice if you’re building anything relating to money, users, or permissions (i.e. everyone)…

- Slow down just a bit
- Reviews, testing, checks & balances are your friends
- Don’t confuse “it works” with “it’s safe”

Speed is a weapon. Unreviewed speed is a liability.

WILD (but unsurprising) stat of the day: 56% of chief execs report NEITHER increased revenue NOR lower costs from AI over the past 12 months.

(from PwC’s 29th Global CEO Survey)

I see it everyday… companies adopting AI because someone above them said, “We need to be doing something with AI.”

But friends… that’s not a business strategy!

AI only works when it’s the best tool for a clearly defined job:

- Reduce time spent on X
- Increase output of Y
- Improve quality or consistency of Z

Track & measure.

If good results → keep 
If bad result/no improvement → dump

That’s it!

PS: Just for clarity… 30% of chief execs reported higher revenues linked to AI over the past year and 26% cited lower costs. So if you’re doing it right… more power to you! I’m just seeing a lot of people NOT implementing AI right - data seems to back me up.

This is a cool moment for me (10+ years in the making). I’m really excited to share that MetaCTO is now a Premium Member of the Orange County OC Startup Council.

I’ve been going to OCSC events for over a decade… first as a founder trying to figure everything out.

And now on the other side helping startups build, launch, and scale.

Back then, OCSC was where I learned about local investor networks, funding trends, and how to pitch without completely embarrassing myself. It was where I met other founders dealing with the same problems I had.

Fast forward to today…

MetaCTO now supports startups full-time. So becoming a Premium Member felt like the right way to give back to a community that gave me so much early on!

Some of my favorite partners and customers are local. I love being able to meet in person, do on-sites, and occasionally end a strategy session with a beer at Salty Bear Brewing.

If you’re building something in Orange County, I can’t recommend OCSC enough.

And if you’re heading to an event soon, come say hi! Happy to meet.

At the risk of sounding like a grumpy old timer…

I miss when software engineers openly cared about the craft.

Not just shipping tickets or hitting sprint goals… but genuinely caring about how things were built. Debating tradeoffs. Obsessing over edge cases. Wanting to make something cleaner or faster even if no user would ever notice.

But I don’t think that mindset disappeared because engineers got worse or lazier. I think the incentives changed.

As a founder now, I see it from the other side.

The industry trained people to optimize for survival: pass interviews, don’t stand out too much, don’t set a pace you’ll be expected to maintain, protect your energy because layoffs are always lurking. 

Add AI and the pressure to “just ship” gets even stronger.

The tension I wrestle with as a boss is “How do I build a team that ships fast AND still cares about quality, learning, and long-term thinking?”

I’m actively trying to solve this inside my own company:

- Rewarding engineers for good decisions (not just output)
- Making time & space for refactors, cleanup, and tech debt
- Being explicit that “working smarter” and “working more” are not the same thing

(I think my team appreciates the last one most)

I genuinely don’t want a team that’s so optimized for productivity and safety that curiosity disappears.

My thinking is if engineers only ever optimize for short-term output, they don’t grow. And if leaders only ever reward output, they shouldn’t be surprised when actually creativity & craftsmanship suffers & fades away.

I don’t have a perfect answer yet. I’m experimenting in real time.

I’m curious: Does anyone know companies zigging (no-AI) while everyone else is zagging (AI)?

Genuine question.

Everywhere I look right now, teams are going all-in on AI… AI in the product, AI in the workflow, AI writing the code, AI reviewing the code, AI deciding what to build next.

But I keep wondering if there are companies intentionally doing the opposite.

I mean teams that are making a deliberate, principled bet on no AI (or extremely constrained AI) in engineering or product.

An interesting thought I’ve been having…

If everyone is outsourcing thinking, judgment, and implementation at the same time, is there upside to being the team that doesn’t? The one that forces a deeper understanding, hard-earned intuition, and human originality/creativity > prompt mastery?

Are there companies doing this on purpose? 

Or is “AI everywhere” already so baked in that opting out is impossible?

Would love real examples (or strong discussion/arguments 😉 ) in the comments!

Most companies don’t use AI…

They ‘perform’ AI.

My MetaCTO co-founder Garrett Fritz was on the 21st Century Entrepreneurship Podcast recently and made a very good point…

AI theater is real.

It looks like this… executives approve AI budgets… then their teams are told to “use AI”…. so they spend thousands (sometimes tens of thousands) a month on AI tools.

And yet… no measurable increase in output.

Because nothing else changed. No ownership or process changes or real incentives.

At MetaCTO, we think of AI as an operating system problem. If AI solves a problem and we can measure the success… we keep it. If not… we dump it.

Pretty simple (and effective).

Bad subscription infrastructure is an app killer.

That’s why we partnered with RevenueCat!

Garrett Fritz and I have helped companies build A LOT of mobile apps at MetaCTO. And one pattern shows up over and over again: the product is solid, users like it…

But subscriptions = the bottleneck

Failed receipts, broken trials, App Store rejections, analytics you don’t trust, teams duct-taping billing logic together. I could go on…

What some companies forget is that subscription infrastructure is part of a growth strategy (just like marketing or sales). If it’s shaky, everything else suffers.

That’s why we're so excited to partner with RevenueCat.

They solve the unsexy (but critical) stuff: billing edge cases, renewals, trials, receipts, integrations, and clean data. It means our clients launch faster, pass reviews faster, and don’t have to rebuild their paywall 6 months in. 

Wins all around!

If you’re building a subscription app and want infrastructure that scales instead of fighting you, happy to chat.

I was told I was CRAZY for investing in SEO in 2025. But Jan 28th… we had 277 organic visits. So… maybe CRAZY smart?!

Jan. 2025: ~1 organic visit per day (on a good day)
Jan. 2026: averaged 100+ organic visits/day (277 peak)

Best part… these thousands of monthly visitors are people actively searching for mobile apps, software dev, and MVP development.

Here’s what’s working for us in the Age of AI:

- We treat SEO as a long-term investment

We started in December 2024 fully expecting nothing to happen for months. SEO only works if you’re patient enough to let compounding do its thing.

- We align SEO with how people use AI now

People don’t just “Google keywords” anymore. They ask questions. They research deeply. They bounce between search engines and AI tools. So we optimized for intent (not keyword stuffing).

- We use a simple formula

a. AI-driven programmatic SEO for volume
b. A small, consistent stream of human-written content for trust
c. A strong, performant website (server-side rendered, fast, accessible, mobile-optimized, clean structure)

That’s it.

This is also exactly how we build sites for clients now. When MetaCTO builds a website, we partner with strong SEO + content teams to bake SEO in from Day 1.

SEO didn’t die. Low-effort SEO did.

Happy to share what we’ve learned if you’re thinking about this for your own product!

I hate seeing #ad. 99% of the time, I won’t use that app / product / service / whatever.

My brain literally goes: “nope, skip.” I know I’m not alone.

Creators hate it too.

They know that if they’re pluggin something and it feels forced… pitchy… salesy, their trust with their audience (built from years of hard work!) is gone.

So what does work? No #ads. But advocates.

I’ve seen high-performing apps avoid traditional #ad entirely by doing rev-share partnerships with creators. The trick is to actually make creators LOVE the product.

The worst way to do this is approaching a creator and saying… “Here’s a script. Post 3 times this month say these bullet points word-for-word.”

Instead, I advise companies to:

- Give creators real access to the app (full premium, not a demo)
- Let them decide how & when it shows up in content
- Pay them when it works (rev-share)

This works 100x better because people don’t necessarily hate marketing. They hate being sold BS from people who obviously don’t care.

I see a lot of companies try to build apps in crowded spaces (fitness, fintech, you name it). Most don’t survive Year 1.

One advantage we have at MetaCTO is visibility.

From 20+ years of experience & partnering with companies like RevenueCat… I’ve had a front-row seat to how apps survive Year 1: what they change, what they rip out, and what they rebuild when things aren’t working.

Take Ladder.

On the surface, they’re “just another fitness app” (one of the most crowded categories on the App Store). But what made them work wasn’t their initial idea. It was how they were willing to reinvent themselves.

- Pricing plateaued early:

They tore up their monetization strategy and rebuilt it based on how users valued the product.

- Growth stalled:

They stopped listening to investors and ONLY listened to members instead. Their users didn’t want more workout videos… they wanted a motivation app. Ladder complied. (I believe this is why they became the highest earning fitness app in 2025)

- Engagement lagged:

They doubled down on brand, community, and meeting users where they already were (TikTok, Youtube, in-person community meet-ups).

20+ years has shown me that every successful app has been unafraid to reinvent themselves.

Please don’t be the company thats clings to the original idea and call it “vision.” 99 times out of 100, the first version is NOT the finished product!

If you hear: “Let’s appoint an AI champion.”

👆 99.999% fail rate.

Okay, let’s assume the best… the person is smart, motivated, capable.

But if they’re handed this golden title and responsibility… but don’t have AUTHORITY to actually do/change/update anything… well, it’s doomed.

Because what happens is pretty predictable. The “AI champion” starts making suggestions, teams push back, and suddenly every decision turns into a debate. They’re stuck convincing, not deciding. Over time, progress slows, politics creep in, and eventually everyone wonders why “AI isn’t working for us.”

And your ‘AI champion’ has now become your ‘AI fall guy’.

Eventually decision-makers assume the problem is culture, training, or effort. But it’s not. It’s ownership. If AI is meant to be strategic, it can’t be optional.

If your ‘AI Champion’ doesn’t have real backing from leadership, you don’t have AI leadership at all. You’ve got a future punching bag.

Authority first.
Adoption comes after.

Code To Cloud

Maintain a centralized platform engineering hub, providing resources, IaC templates, blueprints, and operational guides to support scalable, secure, and efficient cloud-native platforms

Ensure clean repository structure, version control, and up-to-date documentation

Manage and troubleshoot Docker environments for CI/CD, local development, and testing, applying DevOps best practices to support platform reliability

Software Engineer

Shopify

Contributed to backend logic and UI components, shipping features that supported millions of merchants and improved platform reliability

Reviewed and validated AI-assisted code, performing logic checks and tests that increased unit test coverage by 10% and ensured deployment reliability

Explored peers’ code and internal systems to deepen understanding of large-scale backend and frontend development and collaborative workflows

Software Engineering Intern

100Devs

Developed modern and accessible full-stack web applications using TypeScript/JavaScript, React, Node.js, and MongoDB, leading to a 10% reduction in bounce rate

Implemented semantic structure for web applications, resulting in Google Lighthouse scores of 85+ and enhanced website performance

Enhanced development cycles by 10% per project utilizing agile project management methodologies, particularly Scrum

Managed code repositories using Git and GitHub, ensuring version control, code integrity, and a streamlined development process

Integrated third-party APIs and services into web applications, enhancing functionality and providing users with seamless, feature-rich experiences

Automated deployment processes with CI/CD pipelines using GitHub Actions, reducing manual intervention and ensuring smoother, more efficient development workflows

Software Developer

Crystal Int'l (Group) Inc. & Crystal Claire Cosmetics Inc.

Streamlined 2D and 3D drawing review and validation processes, resulting in a 20% reduction in errors and improved business efficiency

Generated custom 3D models using innovative 3D printing techniques, leading to a 10% cost reduction and faster product development by 20%

Facilitated the development process by preparing 3D-printed prototypes for demonstration purposes, resulting in a 15% reduction in the overall development cycle

Effectively communicated technical specifications and feasibility adjustments between clients and manufacturers, streamlining the development process and improving project timelines by 10%

Leveraged collaboration and communication skills with engineers, clients, and manufacturers to align project goals and resolve technical challenges, improving workflow efficiency and reducing production delays

Packaging Development Engineer

Conducted 350+ industrial tests using specialized equipment and adhering to ISTA, ASTM, and ISO standards

Improved pass-test rate by 20% by creating and maintaining process documentation and escalation reports

Generated detailed test reports with photos of damaged areas and recommended potential improvements, resulting in a 10% decrease in retest rates

Coordinated with engineers on 10+ projects, providing technical support and collaborating to resolve challenges, ensuring successful test outcomes and enhancing overall project efficiency

Packaging Test Engineer

🌊 🖥 Software Developer | TypeScript | React | Next.js | Redux | Node.js | MongoDB | ✍ Blogger

Software Developer and SAIT student building production‑ready full‑stack applications, architecting databases, and designing cloud‑ready systems. Skilled in React, TypeScript, Python, Node.js, PostgreSQL, MongoDB, Docker, and CI/CD.

Interned at Shopify, contributing to backend services and UI features used by millions of merchants. Enhanced AI‑generated code for reliability and test coverage, and gained hands‑on experience working across a large‑scale, collaborative codebase.

Previously, as a Packaging Engineer, I developed valuable soft skills in problem-solving, cross-functional collaboration, communication, and time management. These experiences have helped me deliver efficient solutions under tight deadlines, whether optimizing code or workflows.

🌱 Expanding My Skills:
- Cloud & Infrastructure: Learning Azure fundamentals and infrastructure concepts.
- DevOps & Automation: Exploring Docker, GitHub Actions, Ansible, and CI/CD pipelines.
- Systems Programming: Building proficiency in Go (Golang) for backend and tooling.
- Infrastructure as Code (Planned): Looking to explore Terraform and other Infrastructure as Code tools.

🔗 Links & Work:
GitHub: https://github.com/victoriacheng15?tab=repositories
BioHub: https://victoriacheng15.github.io/bioHub/

--------------------------------------------------------------------------------------------

Languages: TypeScript, JavaScript, Python, Ruby (learning), Go (learning)
Frontend Development: React, Next.js, Redux, React Native, Tailwind CSS, SASS
Backend Development: Node.js, Flask, Ruby on Rails,  MongoDB, PostgreSQL
Testing: Cypress, Jest, Testing Library, Pytest
DevOps & CI/CD: GitHub Actions, Docker, Linux, Bash, Ansible, Azure, CI/CD Pipelines

taplio

The Battle

Christopher Fitkin

Hsiu-Chia Cheng

Followers

Posts

Engagements

Likes

Performance ScoreChristopher Fitkin's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

Christopher Fitkin

Hsiu-Chia Cheng

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);}Christopher Fitkin's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScoreChristopher Fitkin's LinkedIn Performance