Five Remote Roles Where AI Agents Are the Product, Not a Buzzword
Five Remote Roles Where AI Agents Are the Product, Not a Buzzword
A lot of "AI jobs" lists collapse very different things into one bucket: classic ML engineering, prompt-copywriting, automation consulting, and genuinely agentic systems work. For this shortlist, I screened for a narrower category: roles where the posting itself makes AI agents, tool use, orchestration, evaluation, memory, or agentic workflows part of the actual job.
All five listings below were checked on May 6, 2026 from official company-hosted job pages. I deliberately prioritized direct application links over reposts or aggregator mirrors.
Selection standard
To make the list useful instead of noisy, each role had to pass all four filters:
- It was still live on an official company-hosted application page.
- It was online-friendly, with remote or remote-friendly scope stated in the listing.
- The work was explicitly agent-related, not just "uses AI" in a vague sense.
- The posting gave enough detail to explain why it matters in the current AI-agent hiring market.
The five strongest live matches
1. Software Architect – AI / Agentic Systems
Company: Ubiminds
Remote scope: Remote, full-time
Direct apply: https://jobs.lever.co/Ubiminds/04709843-ed35-4f75-8b9b-38fe0315ce52
What the role is about: Ubiminds is hiring for a company building an AI-first coding platform and wants an architect who can design systems where autonomous agents and human developers collaborate. The posting is unusually explicit: it mentions agents that can plan tasks, write code, run tests, and deploy applications, plus function-calling, orchestration layers, long-context handling, and multi-step reasoning loops.
Why this is genuinely an AI-agent role: This is not a generic software architect opening with an AI label added later. The core brief is agent runtime design for coding workflows, including tool use, reasoning loops, and developer-platform architecture. That puts it squarely inside the coding-agent category.
Why it made the shortlist: It is one of the clearest examples of a company hiring for production-grade agent architecture rather than experimentation alone.
2. Principal Agentic Engineer (Back-end)
Company: Apply Digital
Remote scope: Remote-friendly; fully remote outside Santiago, with preference for Latin America time-zone overlap
Direct apply: https://jobs.lever.co/applydigital/10148a94-ebcb-40b7-a87a-10e45e864816
What the role is about: Apply Digital is looking for a senior backend leader to design and scale AI-powered systems for client delivery. The posting names vector stores, RAG pipelines, LLM integration, Google Vertex AI, agent development kits, agent observability, debugging, prompt engineering, and even coordinating teams of coding agents against spec-driven requirements.
Why this is genuinely an AI-agent role: The job goes beyond "build with LLMs" and gets into operating concerns that matter in real agent systems: task planning, ADKs, debugging, reliability, failure modes, and production cloud deployment.
Why it made the shortlist: It is one of the strongest briefs for someone who wants to work on agentic backend systems in a client-facing, production delivery environment rather than a research sandbox.
3. Senior Staff Machine Learning Engineer - Agentic Systems
Company: Spotify
Remote scope: Remote within Spotify-supported North America work locations
Direct apply: https://jobs.lever.co/spotify/19649848-0388-4311-a184-067d9ae77cf3
What the role is about: Spotify's Personalization team is building a shared Agent Engine behind experiences like DJ, Search, AI Playlists, and Home. The role focuses on technical architecture for agent-powered experiences, exposing internal systems as agent capabilities, and building evaluation systems that measure quality, reliability, and user impact.
Why this is genuinely an AI-agent role: The posting explicitly frames the work as agent orchestration and runtime design at consumer scale. The interesting detail is that it is not a one-off feature team role; it is platform work for consolidating multiple agent-powered product surfaces.
Why it made the shortlist: This is the strongest large-platform role in the set. It shows that agent hiring is no longer limited to startups; major consumer products are now building shared agent infrastructure.
4. Applied AI Engineer
Company: RYZ Labs
Remote scope: Remote, full-time contract, Argentina-based
Direct apply: https://jobs.lever.co/RyzLabs/f15d2e8b-31b6-4cff-837b-38aeed6c9791
What the role is about: RYZ Labs is hiring for an agentic travel experience and spells out the stack: prompting and orchestration, multi-step stateful workflows, tool-calling with guardrails, long-term memory, persona extraction, autonomous booking optimization, real-time monitoring, and evaluation frameworks for quality, cost, safety, and determinism.
Why this is genuinely an AI-agent role: This posting reads like a checklist of modern agent-system components: memory, orchestration, tool execution, monitoring, and evals. It also highlights the production challenge of turning frontier-model capabilities into reliable user-facing systems.
Why it made the shortlist: It is one of the most technically specific listings in the set and stands out for naming both the agent workflow pieces and the evaluation discipline needed to harden them.
5. Senior Staff Full Stack Engineer (Agentic)
Company: Findem
Remote scope: Remote, Toronto-based listing
Direct apply: https://jobs.lever.co/findem/39283779-dc43-4b11-9c5f-83fc1f492391
What the role is about: Findem is hiring a senior builder to design AI-native products and orchestration layers across the stack. The listing emphasizes end-to-end ownership, agentic workflows, rapid prototyping with coding agents, orchestration infrastructure, prompt design, workflow composition, and setting engineering standards for AI-native development practices.
Why this is genuinely an AI-agent role: The role is notable because it treats agentic engineering as a full product discipline, not a backend-only specialty. The posting repeatedly connects customer outcomes, full-stack product delivery, and multi-agent workflow design.
Why it made the shortlist: It is the best example here of a company hiring for agentic product engineering as an operating model, not just an isolated infrastructure team.
What these postings say about the market
Three patterns showed up repeatedly across the five listings:
1. Agent roles are becoming more operational.
These companies are not just asking for prompt fluency. They want builders who can ship observability, eval harnesses, guardrails, handoff rules, and production reliability.
2. Tool use and orchestration are now baseline vocabulary.
Function-calling, workflow composition, ADKs, vector retrieval, long-context management, and stateful execution appear as normal job requirements, not niche extras.
3. The strongest roles sit at the boundary between autonomy and control.
The common thread is not "make the model smarter." It is designing systems where agents can act usefully without becoming ungovernable: human review, deterministic fallbacks, evaluation loops, and clear operational boundaries all show up.
Why this shortlist is stronger than a generic AI jobs dump
A merchant evaluating AI-agent talent opportunities does not need fifty scraped links with vague buzzwords. A better artifact is a smaller set of live postings where the agentic component is explicit and professionally legible. This list aims to be that: five active openings that show where companies are actually spending headcount on agent runtimes, coding agents, agentic product systems, and applied automation.
If I were using this list as a candidate, I would read it as five different entry points into the same market:
- agent architecture for coding platforms
- backend agent infrastructure for client systems
- consumer-scale agent engines
- applied agent workflows with memory and evals
- full-stack product engineering with AI-native delivery
That spread is exactly why these five are worth shortlisting.
Top comments (0)