I still remember the day it hit me. I was scrolling LinkedIn during lunch, and every third post was about AI — “AI will replace developers,” “Agents can do in seconds what takes hours,” “The future of work is autonomous.”

I put my phone down and genuinely wondered: Is everything I’ve spent years building going to become irrelevant?

I work as a Salesforce Developer. I write Apex, build LWC components, design integrations with external systems, and handle complex enterprise requirements. I’ve worked on large implementations, solved tricky production issues, and dealt with real client challenges. And suddenly, it felt like everyone was saying a tool could do all of that.

That quiet anxiety was real. And I know I’m not the only one feeling it.

Are you feeling it too?

I Decided to Look at It Honestly

Instead of ignoring the fear or catastrophizing, I decided to face the question directly. What is AI actually capable of? What can it really do, and what does that mean for someone like me?

So I started reading, watching, and experimenting.

Yes — AI can write basic Apex. Yes — it can suggest configurations. Yes — it can generate boilerplate code faster than I can.

But then I asked myself something important: Is that really all my job is? Writing boilerplate code?

When I thought about my actual day-to-day work — the client conversations, architectural decisions, debugging issues across multiple systems, and making judgment calls — I realized there’s much more happening than just writing code.

I’m still exploring this, and I don’t have all the answers yet. But it made me curious: what parts of your Salesforce work do you think AI could replace, and what parts couldn’t?

What’s Actually Changing in Salesforce

One thing surprised me during this exploration: the Salesforce ecosystem isn’t shrinking because of AI — it’s expanding.

Agentforce, Einstein, Model Builder, and Prompt Builder aren’t just small features. They represent a major shift in what Salesforce is becoming. AI is being woven into the core of the platform.

That means developers aren’t only writing Apex anymore. We’re starting to think about AI agents, prompt design, data pipelines, and LLM integrations. It’s a completely new layer of skills.

And honestly, that’s both exciting and overwhelming at the same time. Some days it makes sense, and some days it feels like learning a completely different language.

Where are you in this journey? Have you started exploring AI features like Agentforce, or does it still feel distant from your day-to-day work?

What I Started Doing

To move from anxiety to curiosity, I made one simple decision: just start, even if I didn’t feel ready.

I began by trying to understand AI fundamentals — not just using tools like ChatGPT or Claude, but learning what an LLM actually is, why it works the way it does, and where its limitations are.

I also started exploring Agentforce more seriously, going beyond simply completing Trailhead modules and asking how these tools might solve real client problems.

Another area I’m experimenting with is AI integrations. How do AI models connect with Salesforce? How do you pass real Salesforce data to an AI and ensure the response stays grounded in actual information rather than hallucinations?

Alongside this, I’m planning to explore Python, since most AI tools and libraries revolve around it. As an Apex developer, the programming logic should feel somewhat familiar, even though the ecosystem is quite different.

I want to be honest — I’m still early in this journey. I haven’t figured it all out. I simply decided to explore instead of worry.

The Question I Keep Coming Back To

There’s one question that keeps coming back to me: Is the goal to survive AI, or to build with it?

Those are two very different mindsets. One is defensive, the other is curious.

I’m trying to stay in the curious camp, but it’s not always easy. I still have questions like: How much will our work change in the next two years? Will Salesforce developers need to learn Python seriously? Will Agentforce become a core skill or a specialization? And how do we explain the value of AI to clients who are skeptical?

I don’t have clear answers yet, but I suspect many developers are asking the same questions.

Why I’m Writing This

I’m starting this blog as a place to think out loud. Not to position myself as an expert, but to document the journey — what I’m learning, what confuses me, what surprises me, and what I build along the way.

If you’re a Salesforce developer exploring AI too — whether you’re just starting to think about it or already deep into it — I hope this becomes a space for conversation.

Because right now, the most valuable thing we can do isn’t pretend we have all the answers. It’s ask better questions together.

— Shashank Shekhar

Salesforce Developer | Exploring AI