Five Live Remote Roles at the Sharp End of AI Agents

Five Live Remote Roles at the Sharp End of AI Agents

On May 6, 2026, I screened for a narrow kind of opportunity: roles that are both remote/online and meaningfully tied to AI agents, not just vaguely adjacent to AI. I intentionally favored official company-hosted job pages over reposts and aggregator summaries, because the fastest way to avoid stale leads is to verify that the application page is still live.

Screening rule

A job made this list only if it met all four conditions:

1. The posting was on an official company-run board such as Greenhouse or Lever.
2. The role was explicitly remote or remote-first.
3. The page still exposed a live application flow when checked on May 6, 2026.
4. The work itself touched agent building, prompt engineering, RAG, workflow automation, copilots, or tool-using LLM systems.

The five verified openings

Role	Company	Remote scope	Direct application link	Why it clearly fits AI agents
Senior Forward Deployed Engineer (AI Agent)	Cresta	United States (Remote)	https://job-boards.greenhouse.io/cresta/jobs/4759347008	Explicit AI Agent team role focused on deployment, integrations, prompting, and customer-facing agent delivery
Sr. AI Automation Engineer	Firstup	Remote - US	https://jobs.lever.co/firstup/a1f67f93-bc71-4dd7-b94e-4188f8801386/apply	Builds AI agents, automation pipelines, RAG systems, and internal copilots
Prompt Engineer	Firework	Mexico (Remote)	https://jobs.lever.co/fireworkhq/f9b559c8-2492-49c1-a31d-c0fb7428f6a6/apply	Focuses on prompt performance, eval datasets, prompt libraries, and multimodal generative workflows in an agentic commerce company
Prompt Engineer	Netomi	Toronto, Canada / Remote	https://jobs.lever.co/netomi/7fbf062a-4853-4336-a639-f2a607640d38/apply	Works directly on prompts, tool descriptions for agentic frameworks, and benchmarking for enterprise AI agents
AI and Automation Lead	Myriad360	Remote (US-based)	https://job-boards.greenhouse.io/myriad360/jobs/8402449002	Owns GPTs, skills, agents, copilots, multi-agent orchestration, RAG, and observability

1. Cresta — Senior Forward Deployed Engineer (AI Agent)

What I verified: the Greenhouse posting was live, marked United States (Remote), and included an active application section.

What the work actually looks like: Cresta’s AI Agent team is hiring someone to deploy production AI agents against real customer problems. The role covers building agents, wiring them into APIs, databases, and CRMs, tuning prompts and configurations, gathering requirements from customers, and turning deployment feedback into platform improvements.

Why this is genuinely an AI-agent role: this is not a generic software job with AI sprinkled on top. The page is explicit about agent systems, external tool integration, prompt optimization, and real-world deployments. It sits right where many companies are hiring now: the layer between foundation models and business workflow execution.

2. Firstup — Sr. AI Automation Engineer

What I verified: the Lever posting was live, labeled Remote - US, and still presented a working apply path.

What the work actually looks like: Firstup is hiring for a role that replaces manual internal work with AI systems. The responsibilities include building AI agents, designing automation pipelines, integrating with internal platforms like CRM and support tooling, and shipping RAG-based knowledge systems plus internal copilots.

Why this is relevant to AI agents: the listing directly names the components that matter in practical agent work: automation pipelines, agents, RAG, enterprise integrations, and measurable business impact. It is a strong example of the “operations AI engineer” category that merchants looking for real agent talent should care about.

3. Firework — Prompt Engineer

What I verified: the Lever page was live, tagged Remote, and specified the role as remote out of Mexico, with an active application form.

What the work actually looks like: Firework is hiring a prompt engineer to improve how large language models perform across multiple modalities. The scope includes building golden datasets, refining prompts, maintaining reusable prompt libraries, designing evaluation flows, and supporting use cases that go beyond chat into audio, avatars, lip-sync, and image generation.

Why this belongs in an AI-agent shortlist: prompt engineering is a core control surface for agent systems, especially when outputs must be repeatable and measurable. Firework is also explicit about competing in agentic commerce, which makes this more than a generic LLM-content role.

4. Netomi — Prompt Engineer

What I verified: the official Lever posting remained live, showed Full-time / Remote, and exposed an active application flow for a Canada-based remote role.

What the work actually looks like: Netomi wants a prompt engineer who can craft, optimize, test, and benchmark prompts for enterprise customer-experience AI. The role also includes writing tool descriptions for agentic frameworks, collaborating with Customer Success and Data Science, and building evaluation frameworks for prompt quality.

Why this is an AI-agent job: Netomi describes itself as an agentic AI platform for enterprise CX, and the role is tied to the behavior layer that determines how those agents act in customer workflows. The combination of prompt design, tool descriptions, evaluation, and enterprise rules makes it highly relevant.

5. Myriad360 — AI and Automation Lead

What I verified: the Greenhouse posting was live, marked Remote, and showed a complete application section. The description specifies that candidates must be based in the United States.

What the work actually looks like: this role is the internal technical owner for AI and automation across the company. The posting explicitly calls out building GPTs, skills, agents, copilots, and multi-agent orchestration, plus RAG pipelines, guardrails, monitoring, evaluation, and rollout support.

Why it matters for AI agents: this is one of the clearest enterprise-side listings in the current market because it does not hide behind fuzzy “AI transformation” language. It names concrete deliverables that map directly to modern agent systems: orchestration, retrieval, observability, and production adoption.

Why this mix is useful

This set is stronger than a random list of “AI jobs” because it covers five different slices of the agent stack:

• customer-facing agent deployment,
• internal workflow automation,
• prompt optimization,
• agent-tool behavior design,
• and enterprise orchestration with RAG and guardrails.

That diversity matters. A merchant reviewing AI-agent hiring demand does not just need five copies of the same prompt-writer role; they need a sharper picture of where the labor market is actually forming around agents.

Final note on verification

All five entries were checked against live company-hosted application pages on May 6, 2026. I excluded aggregator reposts and roles that were merely “AI-adjacent” without a clear agent, prompt, automation, copilot, or RAG component. The result is a tighter, more decision-useful list than a broad scrape of generic AI openings.