25 verified roles · one editorial · every Monday, free.
Forward Deployed Engineer · Applied AI Engineer · AI Automation · Agent Operations.
Same work. Different titles. One newsletter.
Independent · no VC · no paid placement
// Section 02 — Definition
On April 20, on the 20VC podcast, Aaron Levie predicted 500K to 1M new jobs for "agent operators" — the people who deploy AI inside marketing, legal, and ops teams.
Three days later, working contractors confirmed they're already doing it. Companies are paying $200K–$320K for senior roles under titles like Forward Deployed Engineer and Applied AI Engineer. The work is the same. The title is in flux.
The Operator covers the work, not the label.
Read the full breakdown →// Section 03 — Title war
After reading 50+ open job postings this week, the same role appears under at least seven names — sometimes at the same company. Compensation is real across all of them.
| Title | Where it shows up | Pay band seen |
|---|---|---|
| Forward Deployed Engineer | OpenAI, Anthropic, Stripe, Palantir, Databricks, Modal, Cursor, Sierra | $200K – $390K |
| Applied AI Engineer | Anthropic, Ramp, OpenAI, WorkOS, NVIDIA, LangChain | $155K – $385K |
| AI Agent Engineer | Sierra, Hebbia, Modal, Relevance AI, Mercor | $230K – $390K |
| AI Automation Engineer | OpenAI (internal), Vercel, mid-market SaaS | $120K – $490K |
| Agent DX Engineer | WorkOS, Modal | $175K – $275K |
| AI Solutions Architect | Stripe (post-sales flavored), enterprise integrators | $ varies |
| Integrations / AI Workflow Engineer | Hebbia, vertical AI cos | $ varies |
// Section 04 — This week
Embed with Anthropic's largest enterprise customers and ship working Claude deployments end-to-end — skills, MCP servers, evals against their data. ~30% on-site travel. Six of the eight Operator skills explicitly listed in the JD.
Ships production GPT deployments inside customer environments — pick the model, wire it into their tools, own the outcome.
Builds customer-facing agent infrastructure at the Bret Taylor / Clay Bavor agent co — model selection, tool-wiring, production ownership.
Builds the agent features Ramp ships to customers — tool-use, evals, production agent loops. Highest comp band in the bucket.
Deploys Perplexity Enterprise Pro inside Fortune-500s — builds custom connectors against their data sources, defines skills against their schemas, designs evals that match their internal benchmarks. MTS title means flatter org, broader scope than typical FDE.
Core agent-platform engineering — owns the orchestration layer, skill registry, and tool-use runtime every Sierra customer's agent runs on. Less on-site than FDE, more systems design, same pricing band. For Operators who'd rather build the rails than ride them.
Internal FDE — deploys AI into Ramp's own workflows (finance ops, customer success), owns outcomes end-to-end.
Explicitly "Agent DX" — builds the developer surface for agents running on Modal. Rare title.
Deploys Cursor into enterprise dev orgs — integrates with their codebases (often monorepos with custom build systems), their CI, their internal tools, their security review workflows. The job is making Cursor not just usable but governed inside a Fortune-500 engineering org. Equity-heavy comp.
Operator role inside OpenAI itself — builds the Codex coding agent's tool-use and skill system. Equity on top of cash band.
15 more roles in this week's issue. Subscribe to receive Monday →
// Section 05 — What you get
A short brief with 20–30 verified roles, one editorial, and one signal worth your time. No filler. No sponsored placements.
Salary ranges. Location splits. Title trends. We track which titles are winning volume and which are losing it.
Monthly interviews with working Operators — what they actually do, how they got hired, what they wish they'd known.
// Section 06 — For employers
Founding-period job posts are free through May 31. After that, $299/role. Listings reach 1,000+ Operators across Forward Deployed, Applied AI, and Automation tracks. Posted Mondays, archived as SEO-indexed pages forever.
Submit a role