From SDR to GTM Engineer: The Career Shift That's Already Happening

Nobody became an SDR because they loved researching companies in six tabs.

They took the role because they believed something: the right message to the right person at the right moment changes a business's trajectory. That belief is still true. What's changing is what it takes to act on it.

Drew Bredvick is Director of GTM Engineering at Vercel. His team built an inbound lead agent in a weekend. They went from 10 SDRs handling inbound to one. Saved $2M+. 32x ROI. The agent does lead research, enrichment, qualification, email drafting — all of it — before a human touches the lead.

The remaining SDR reviews the AI's work and presses send.

As Drew put it: "It's like we gave them the controller to a video game instead of making them play the boring parts."

---

Research, enrichment, qualification, email drafting, follow-up — agents handle all five now, continuously, at scale, without getting tired or demoralized by the no-replies.

This is already here. Vercel is one data point; there are hundreds more.

The SDR function is disrupted. The question is what the people who understand that function do next.

---

The answer is to move up the leverage stack, not just learn to code.

Drew's definition: "GTM Engineering is pattern recognition turned into automation: find what works, then build agents that do it better every time."

The most valuable thing an SDR carries is their understanding of the buyer — what signals matter, what objections come up, what language lands. Not the sequencing tool. Not the call cadence.

That understanding is what the agent needs to be useful. Someone who's spent two years as an SDR knows more about building a useful lead qualification agent than most engineers do — they've lived inside the problem and know where the rubric breaks.

---

You don't need a CS degree to do this work.

There are two paths depending on your technical comfort. More technical: Cursor or Claude Code. You describe what you want to build in plain language, the AI writes the code, you ship it. Less technical: n8n, Zapier, or Gumloop — canvas-based workflow builders where you connect enrichment APIs, qualification logic, and outreach sequences without touching code. You're still building agents. You're just using a different interface.

Either path leads to the same place. Someone who can automate a GTM workflow end-to-end is valuable in ways that are hard to replace, and that's the only distinction that matters now.

Drew on the bar to clear: "Ship an AI automation that you wrote, deployed yourself, and that runs in production. Most applicants talk about AI. Almost none have gotten their hands dirty."

---

Where to start: ask an LLM to train you.

Pick one workflow you do repeatedly — lead research, email personalization, follow-up sequencing. Tell Claude or ChatGPT you want to automate it and you're new to this. Ask it to walk you through building it step by step using n8n or Zapier or whichever tool fits your level. It will. Follow through and ship the thing.

Drew: "Don't ask for permission. Don't wait until you feel ready. Pick a small project, something low stakes, and ship it this week. The people who break into GTM Engineering don't know more than you. They're just the ones who decided to start."

---

Not every SDR will make this shift. Some will wait and find the role has changed under them. The ones who move early will discover they're extraordinarily good at it — because they understand the domain and can now automate it.

The career arc from here is clear. GTM Engineer to broader agent orchestration to one of the more durable, higher-compensated roles of the next decade. The people who can build systems that think in sales terms are rare, and they're going to be rarer relative to demand, not more common.

The SDR skill set transfers. The SDR job description doesn't.