Code, Media, and Leverage: How I Navigated the 2025 Tech Market
The messy reality of building leverage, surviving 2025, and transforming for the AI era.
We’re a few weeks into 2026 now, and with some distance, it’s easier to see 2025 for what it really was: a correction year.
Post-COVID excess finally unwound. Hiring slowed, budgets tightened, and the bar for technical roles jumped, not just on depth, but on relevance. Being “good at coding” stopped being enough. Companies wanted people who could ship, explain why it mattered, and connect technical decisions to business outcomes from day one.
I don’t claim to have cracked some secret code. I made plenty of wrong turns, second-guessed myself often, and benefited from timing and luck as much as effort. But I did change how I approached the market, and that shift helped me land a Staff-level role going into 2026.
What follows isn’t a blueprint, just the lessons that made the biggest difference for me.
1. Advertise Yourself (Build Your Signal)
There’s a comforting myth in engineering: “Good work speaks for itself.”
In a noisy market, that’s simply false. Good work whispers. Noise wins.
For a long time, I assumed that if I just kept my head down and built solid systems, the right people would eventually notice. In 2025, that assumption stopped holding. If people didn’t know what I was working on, what I was curious about, or how I thought about problems, I effectively didn’t exist.
So I stopped treating my work like it was in permanent stealth mode. I started sharing learnings, partial ideas, architectural tradeoffs, and even things I hadn’t fully figured out yet. I leaned heavily on a framework popularized by Naval Ravikant: Code and Media as permissionless leverage.
Not everything landed. Some posts went nowhere. Some ideas aged poorly. That was uncomfortable but necessary.
The lesson: You might be capable, but value is invisible until it’s communicated. Being visible isn’t arrogance; it’s clarity. When you increase your surface area of luck, the market has something to react to.
2. Learn a New Skill (But Don’t Chase Hype)
“Learn whatever’s hot” is lazy advice. If you chase every trend, you end up broad, shallow, and forgettable.
Instead, I tried to be deliberate about adjacency. I didn’t abandon my background in data engineering; I extended it. I focused on the bridge between where I already had depth and where the market was clearly heading: AI systems, RAG, GraphRAG, real-time architectures, and the unglamorous infrastructure problems underneath them.
In version two of this story, it might sound like a clean pivot from “data plumber” to “AI architect.” In reality, it was messier. I built things that didn’t scale. I over-engineered. I underestimated operational complexity more than once. But I was learning in the direction of leverage.
The lesson: Don’t learn tools for their own sake. Learn skills that compound your existing experience in a new context. Be the bridge, not the tourist.
3. Code + Media = Leverage
Naval Ravikant’s idea that code and media are permissionless leverage sounds abstract until you live it.
Code was still the foundation. Building real systems, especially ones that touched retrieval, orchestration, streaming, and AI infrastructure, gave me credibility. Shipping mattered far more than talking.
But code alone didn’t close the loop.
What surprised me was how often media made the difference, not podcasts or YouTube, just clear communication.
Your resume is media.
Your interview answers are media.
Your blog posts, diagrams, and “About Me” blurbs are media.
I noticed a pattern: my code got me in the room, but my ability to explain why the system was designed a certain way and what tradeoffs I accepted was what moved conversations forward.
That didn’t come naturally. I rewrote explanations repeatedly. I stumbled in interviews. I learned (slowly) how to connect technical decisions to business risk, cost, and reliability.
The lesson: Code builds the value. Media translates it. If you neglect either, you leave leverage on the table.
4. The Recruiter Is an Ally (If You Let Them Be)
Earlier in my career, I treated recruiters like hurdles, something to clear before talking to the “real” team. That mindset cost me opportunities.
In 2025, I changed tactics. I started treating good recruiters as partners. I asked direct questions. I asked for feedback. I listened for what wasn’t written in the job description.
Recruiters often know:
What the hiring manager is worried about
Why the last candidate failed
Which skills are negotiable and which aren’t
They don’t always volunteer that information, but they will if you treat them like collaborators instead of obstacles.
The lesson: Prep with your recruiter, not just for them. An informed ally on the inside is a real advantage.
5. Using AI as a Force Multiplier (Not a Crutch)
One thing I learned quickly: if you’re applying for AI roles and not using AI in your job search, you’re leaving efficiency on the table.
I treated LLMs like a Chief of Staff, not a resume-spamming machine.
I used them to tailor resumes, not fabricate experience.
I used them to synthesize company strategy and competitive landscapes.
I used chat-based mock interviews to stress-test my explanations.
The outputs weren’t perfect. I had to sanity-check everything. But the time savings and clarity were real.
The lesson: AI won’t replace judgment, but it dramatically amplifies it if you already know what “good” looks like.
Looking Ahead: 2026 and the Shift to AI Production
If 2025 was the year of the AI demo, 2026 is shaping up to be the year of AI production.
The novelty is gone. A chatbot that works 80% of the time is no longer impressive. What matters now are the hard, messy problems: Governance, Security, Accuracy, Cost, and operational reliability. We’re moving from the magical phase of AI to the industrial one. The opportunities won’t go to people who can use tools, but to those who can make systems dependable.
I’m still learning. I’m still wrong more often than I’d like. But the direction is clear.
Keep building. Keep shipping. And don’t be afraid to tell your story - clearly, honestly, and without pretending you have it all figured out. The opportunity is out there.
Wow, the part about 'good work whispers' really hit home! It’s so true. What if everyone just kept their heads down? We’d miss out on so much innovation and shared learning. Makes me think how I encourage my students to showcase their projects too. Great read!