“What 1,247 active implementer job postings actually require.”
We parsed the canonical capability taxonomy across every implementer-classified role in our feed. Here's what the market is asking for:
- Claude appears in 64% of postings. This is up from 41% a year ago and reflects both Anthropic's go-to-market acceleration and a clear lab-side preference.
- GPT-4o is at 48% — still ubiquitous but no longer dominant.
- RAG (41%) and Cursor (34%) are the most common adjacent skills. RAG is now baseline.
- Voice agents (27%) is the fastest-growing.
- LangGraph (22%) has eclipsed bare LangChain in new postings.
- n8n (18%) is the dark horse — quietly the canonical tool for non-code-first implementers, especially in mid-market ops.
- Fine-tuning (16%) is more common than you'd think; most postings asking for it are actually asking for "evals + small-model finetune", not GPT-style base-model training.
The takeaway for implementers: Claude + Cursor + RAG + voice + one workflow tool (n8n or LangGraph) covers ~85% of postings. Get fluent on those five and you can answer most JDs.
The takeaway for operators: if your shortlist doesn't show Claude experience, you're sourcing from a stale list.
Get the next one in your inbox.
We don't publish often — when we do, it's because we found something worth sitting with.
- Tool deep-diveVoice agents quietly took over enterprise intake
Three under-the-radar deployments in healthcare, plus the four tools every implementer now needs to know.
- ThesisThe Forward Deployed Decade
Why the role Palantir invented in 2003 is the highest-leverage seat of the next ten years — and the labor-market repricing that's already underway.