I found myself in Brooklyn this week for AI Native DevCon, and if you haven’t been paying attention to this community yet, trust me — you will. What’s emerging here isn’t just a conversation about AI in development. It’s a blueprint for how platform engineering will evolve in an era defined by intelligent agents, AI-generated workflows, and policy-enforced development ecosystems.
The name “AI Native Dev” isn’t metaphorical.
It’s literal.
It’s about building software with AI baked into the foundation, not bolted on after the fact. And the platform engineering angle? That’s the heartbeat of the entire movement — even if not everyone fully realizes it yet.
This isn’t just the future of Dev.
This is the future of platform engineering.
Industry City, Brooklyn: The Right Launchpad
The event was held in Industry City, my first time there. Outside, you could freeze a cup of coffee just by holding it out at arm’s length. Inside? The place was radiating energy — whiteboards filling up, clusters of engineers swapping ideas about agent orchestration patterns, people arguing about how to encode LLM guardrails directly into CI/CD pipelines.
This wasn’t Silicon Valley’s hyper-polished tech theater.
And it wasn’t Manhattan’s enterprise marble-and-steel seriousness.
This was Brooklyn: Independent, creative, gritty, full of builders.
Platform engineering has always been born in places like this — where the tinkerers, the system thinkers, and the SREs who keep the world stitched together gather to invent the next layer of abstraction.
This event felt like that kind of birthplace.
A Community That Looks Very Familiar
Walking through the crowd, I couldn’t shake the déjà vu.
Because the faces, the energy, the attitude — everything about this community — reminded me of the earliest days of DevOps.
And that’s no accident.
One of the original catalysts, Patrick Debois, is actively involved. Anytime Patrick shows up, you know the community is real. DevOps didn’t become DevOps because of marketing. It became DevOps because people like Patrick anchored it in experimentation, sharing and collaboration.
At the center of the movement, you’ll also find Guy “GuyPo” Podjarny, Snyk co-founder, and now building Tessl, his new AI-native venture. Alongside him is Simon Maple, another long-time community-builder with deep developer DNA.
These aren’t tourists.
They’re people who have been in the trenches, scaling security, DevOps, developer experience and platform engineering for over a decade.
And their involvement makes something very clear:
Security and platform engineering aren’t afterthoughts in AI Native Dev. They’re foundational.
Who Showed Up? Platform People.
More than 300 people attended the two-day conference, and they were exactly the types who tend to show up early in movements that matter:
- Platform engineers
- SREs
- DevOps veterans
- Agent architecture nerds
- Security engineers
- Builders who like to get their hands dirty
- People who’ve already built their own AI pipelines “just to see what happens”
Notably absent?
The monetization talkers.
The buzzword chasers.
The “growth at all costs” crowd.
This was still pure.
Still hands-on.
Still about solving real engineering problems.
The Big Shift: 90% of Developers Already Use AI
One stat kept surfacing:
90% of developers already use AI.
But most of that usage is scattered, uneven and happening in the shadows:
- Vibe coding with no guardrail
- Independent agent experiment
- Ad-hoc LLM prompts
- Unvetted tools running in personal dev environments
- Zero governance
- Zero policy
- Zero platform alignment
To a platform engineer, that’s not just messy.
It’s a potential nightmare.
AI Native Dev isn’t trying to restrain AI usage.
It’s trying to operationalize it.
It’s saying:
Let’s stop making AI a sidecar and make it part of the platform.
Agenda Highlights Through the Lens of Platform Engineering
Scanning the agenda at AINativeDev.io/devcon, what struck me was how platform-centric this movement already is — even if not always explicitly labeled that way.
Agent Orchestration Patterns
How do different agents negotiate tasks?
How does the platform supervise them?
How do we prevent infinite loops, dependency deadlocks, or runaway task chains?
LLM Governance Built Into the Platform
Not governance as in “procedures,” but governance as in:
Encoded policies that shape what AI is allowed to do.
Secure AI Usage for Dev & Ops
Security controls as platform primitives.
Guardrails as code.
Context models that agents consume automatically.
AI-Augmented Build, Deploy & Operate Loops
Think GitOps + AIOps + agent assist + validation loops.
Everything becomes faster — but also safer — because the platform is pre-wired with best practices.
Platform Teams as AI Force Multipliers
This is the big idea that kept surfacing:
Platform teams won’t just standardize tools — they’ll standardize AI.
This is the next iteration of Platform Engineering.
Simon Maple on Guardrails and the “Cowboy AI” Problem
One of the clearest statements of the conference came from Simon Maple. Here’s a refined version of his quote:
“To use AI securely in dev and DevOps, we need platform-level best practices — the guardrails, policies, and context we embed directly into the environment. Platform teams can distribute these across every developer workspace, reining in some of the ‘cowboy’ approaches we’re seeing with vibe coding.”
Ultimately, these controls are about LLM guidance: Giving agents the right context for how we build, not dictating what we build.”
— Simon Maple, Head of DevRel & Customer Success, Tessl
This is the whole story.
Platform engineering isn’t here to control developers.
It’s here to control the AI that developers (and operators) increasingly rely on.
The Endgame: Platforms That Power Agentic AI Workforces
Here’s where the movement gets really interesting, especially for PlatformEngineering.com readers:
This isn’t just about helping developers today.
It’s about building platforms for digital workers tomorrow.
The endgame is a world where:
✔ Agentic AI workers can move through the entire SDLC
Based on platform-encoded policies, rules, validations and patterns.
✔ Best practices are baked into the platform
And not written in Confluence pages or spoken in meetings.
✔ Security is transparent, universal, and automatic
Because every agent operates within a policy-defined environment.
✔ Ops becomes event-driven and AI-augmented
With agents triaging incidents, suggesting remediations and deploying fixes.
✔ Platforms become the brains and guardrails
Of the entire engineering ecosystem.
The AI-powered future isn’t going to be defined by tools.
It’s going to be defined by platforms.
Shimmy’s Take
As I stood there in Industry City — watching engineers sketch agent flows on whiteboards, listening to debates about embedding governance into pipelines, seeing the spark in people’s eyes — I felt something familiar.
I saw the early DevOps movement.
The spirit.
The curiosity.
The willingness to break the rules and write new ones.
Then, just as quickly, I saw the next stage of that evolution.
It reminded me of Jon Landau at that Harvard Square show:
“I saw the future of rock ’ n’ roll and its name is Bruce Springsteen.”
Well, I saw a future too — one where Dev, Ops, and Platform Engineering converge under the banner of AI-native design.
I saw the future of platform engineering — and its name is AI Native Dev.