That is the one line I am taking with me after finishing DevRel Uni Cohort 7.
Before this cohort, I had a lot of questions in my head.
Will AI replace DevRel/other roles?
Are docs still important?
Do communities still matter?
What even is my role in all of this?
Five sessions later, I have answers.
And honestly, my whole way of thinking has shifted.
So let me share what I learned, what surprised me, and what I am taking forward.
TL;DR: All 5 Sessions in One Place
Session 1 - Bianca: The core of DevRel (sensemaking, storytelling, distribution, feedback loops) stays the same. But the audience, docs, content, and community are all changing because of AI.
Session 2 - Nader Dabit: Do not lock yourself into one title. Become a technologist. Keep learning, keep building, keep moving. The safest thing in tech is being adaptable.
Session 3 — Patrick Skinner: Speed without systems is chaos. Build a Brainlifting habit — learn daily, document what works, and keep your knowledge system alive.
Session 4 — Hassan El Mghari: Developer Experience is not enough anymore. Agent Experience is the new frontier. Your docs need to be readable by humans AND AI agents.
Session 5 — Francesco Andreolí: DevRel is becoming more important, not less. GEO is the new SEO. Build systems that help both humans and agents grow your ecosystem.
One big theme connects all of them:
DevRel in the age of AI is not about doing less. It is about doing more, with better systems, for a bigger audience that now includes AI agents and LLMs.
> So the real skill is simple: be adaptable.
What Actually Changed for Me
Here is what shifted in my head:
1. My audience is not just developers anymore.
It is developers, founders, students, non-coders, and AI agents too. That changes everything about how I write, teach, and build.
2. Docs are not just for humans.
If ChatGPT or Claude cannot understand my docs, my product loses. Writing for LLMs is now part of the job. GEO (Generative Engine Optimization) is real.
3. I do not need to pick one identity forever.
Nader's line, "become a technologist", — broke something open for me. I was so worried about choosing the "right" path. Now I just want to keep learning and stay adaptable.
4. Speed is not the goal. Systems are.
Patrick's session hit hard. AI can make me 10x faster. But without systems for learning, documenting, and reviewing, I am just creating chaos faster.
5. Trust is still human.
This is the part that gave me peace. Sensemaking is human. Storytelling is human. Community is human. AI cannot replace any of that. It can only scale it.
What Surprised Me Most
Honestly? How much the human part of DevRel still matters.
I thought the whole cohort would be about new tools, new workflows, and new automations.
And yes, we covered all of that.
But every single mentor - Bianca, Nader, Patrick, Hassan, Francesco kept coming back to the same thing:
Taste. Judgment. Trust. Community.
You cannot automate those.
You can only build them by showing up, being honest, and helping people.
That surprised me. In a good way.
Two Sessions That Stuck With Me
Every session was valuable. But two of them keep playing in my head.
Nader's session on navigating a tech career.
Nader did not just give career advice. He gave a way of thinking about a career.
The line that stayed with me was: become a technologist.
Not a frontend developer. Not a DevRel. Not a content creator. A technologist.
Someone who learns what is needed, builds across domains, and adapts as the industry moves.
That reframe took the pressure off picking the "right" path. Because in tech, there is no right path. There is only the next move.
He also talked about how careers are not straight lines. They look messy in the moment. They only make sense looking back.
That gave me permission to experiment without needing a guaranteed outcome at every step.
Hassan's session on building apps and Agent Experience.
Hassan did not just talk about AI. He showed how he actually builds with it, coding agents, demo apps, documentation bots, internal tools, and automations.
It was not a theory. It was real workflows from someone shipping every week.
But the part that stuck was this:
He is not just building apps for users. He is building apps that help people learn, ship, and explore a product faster — with AI inside the workflow.
And the bigger idea, Agent Experience, changed how I think about my own work.
If a developer reads my docs, that is one user.
If an AI agent reads my docs and recommends my product to hundreds of developers, that is leverage.
So now I am not just asking, "Is my doc clear for a human?"
I am also asking, "Can an agent understand it well enough to actually act on it?"
That is a new lens. And it is hard to unsee once you see it.
How AI Changed The Way I Build and Learn
I was already using AI before this cohort.
So the shift was not "I started using AI."
It was "I started thinking about AI differently."
Before, I used AI to help me build.
Now, I also build for AI.
That is a small switch in words. But a big switch in mindset.
A few habits I picked up during the cohort:
I write docs that work for humans first, but I also check if an AI agent can understand them.
I think about how my content shows up in ChatGPT, Claude, and other AI tools, not just Google.
I build systems around my AI workflow, not just speed.
I treat AI as part of my audience, not just my toolkit.
AI does not give me ideas. It helps me move faster on the ideas I already have.
But the bigger unlock from this cohort was realising AI is also a reader, a learner, and a distributor of my work.
That changes everything.
What I Built During the Cohort
This cohort also pushed me to ship.
I worked on DriftGuard, a developer tool that came from a simple frustration:
SDK docs often drift from the actual code.
A docs example works at one point, then the SDK changes, and nobody notices until a developer tries it and it breaks.
That is painful.
So I shipped the first version of DriftGuard to npm.
npm install --save-dev @driftguardjs/cli
The idea is simple.
You point it at your SDK, snapshot once, and commit the baseline. Then every PR can check for broken doc snippets, removed exports, or changed signatures.
If something drifts, DriftGuard can fail the check and show the file and line where the issue happened.
It is still early, but building this during the cohort helped me connect the sessions to a real problem.
Because DevRel is not only about explaining things better.
It is also about reducing friction for developers.
And that is what I want DriftGuard to help with.
What I Am Taking Forward
A few things I am committing to after this cohort:
- Build in public. Even when it feels small. Even when I am not an expert.
- Write for both humans and agents. Every doc, every post, every README.
- Build a learning loop. Learn, build, share, get feedback, repeat.
- Keep my network warm. Not transactional. Just real.
- Stop waiting to feel "ready." Move when the window opens.
- Treat every day like Day 1.
Thank You, DevRel Uni
A huge thank you to Bianca and the entire DevRel Uni team for putting this together.
You did not just teach us frameworks.
You brought in mentors who have actually done the work.
And you created a space where we could ask honest questions and meet honest people.
I met so many incredible builders in this cohort.
People I know I will stay in touch with for a long time.
That alone made the whole thing worth it.
And to every mentor — Bianca, Nader, Patrick, Hassan, and Francesco, thank you for being so generous with your time and your truth.
If You Are Reading This
If you are even a little bit curious about DevRel, AI, or building developer ecosystems, join the next DevRel Uni Cohort.
You will not just learn frameworks.
You will meet your people.
And in the age of AI, that is the most valuable thing you can do.
The fundamentals of DevRel are not changing.
The tools are just getting 10x more powerful.
So learn AI.
Use it.
And become better at helping people build.
See you in the next chapter!
Top comments (0)