About Metacto
At Metacto, we partner with companies across industries to transform how work gets done. We build AI agents and agentic workflows across customer operations, finance, revenue, support, and beyond.
Working at Metacto means tackling a constantly changing mix of ambitious, real-world problems. You might redesign a critical business workflow, connect fragmented systems and data, or build an agent that becomes part of a client’s daily operations. You’ll contribute across the full lifecycle, from understanding the business problem to designing, shipping, evaluating, and continuously improving the solution in production.
We move quickly, give people meaningful ownership, and measure our success by the business outcomes we create. If you want to work at the forefront of applied AI, solve challenging problems across a diverse set of clients, and help define how agentic software is built and deployed, we’d love to hear from you.
Job Overview
We’re looking for an AI Engineer to own agents end to end on our operational AI engagements.
You have shipped production systems and you want to own agents end to end, inside real client environments, with the evals, the deploys, and the on call that come with that. You will carry one agent or workstream from its charter to the day operations depend on it.
Key Responsibilities
- Own an Agent End to End: Take an agent from its charter through build, evals, deployment, and the incidents that follow, including orchestration, prompts, retrieval, tool and function calling, structured outputs, workflow state, and internal API integration.
- Build Applied AI Features: Implement services, workflows, and reusable components for LLM-powered automation, retrieval, tool use, summarization, classification, decision support, and knowledge workflows.
- Support Agentic Workflows and Integrations: Build workflows that take useful action against real systems while staying auditable and controlled, with human review checkpoints, escalation paths, retries, and failure handling at every step that touches production.
- Develop Retrieval and Knowledge Systems: Contribute to RAG and agentic retrieval pipelines over enterprise content and operational data using embeddings, vector databases, hybrid search, reranking, citations, access controls, and freshness strategies.
- Validation Against Real Data: Test the approach against the client’s production data, systems, and failure modes before committing to a build plan, and document what the test showed about feasibility and risk.
- Improve AI Quality, Safety, and Evaluation: Create and maintain evaluation suites, regression tests, prompt and model versioning, trace analysis, guardrails, policy checks, PII handling, and hallucination mitigation for the agents you own.
- Production Engineering and Deployment: Build scalable APIs, services, and event-driven workflows in Python or TypeScript, deploy them with monitoring and rollback paths, and own the incidents on your own work.
- Cloud Delivery and Automation: Ship into client environments using containers, CI/CD, infrastructure as code, secrets management, observability, and runbooks their team can actually operate, inside their access model and their constraints.
- Solve Business Problems with AI: Work with operations, product, data, and engineering stakeholders to understand use cases, prototype solutions, measure outcomes, and move proven capabilities into production.
- Cross-Functional Collaboration: Participate in design reviews, implementation planning, troubleshooting, documentation, and knowledge sharing across technical and non-technical teams, and route scope and commercial questions to the Transformation Lead.
Requirements
- 3+ years shipping software, with at least one LLM or agent system running in production with real users.
- Production Python or TypeScript, and comfort working in whatever stack the client already runs.
- Working depth in retrieval, prompt and context engineering, tool and function calling, structured outputs, and agent orchestration.
- Evals discipline: you can say how you knew it worked, what it did when it failed, and what you changed.
- A deploy you owned end to end, and the incident that followed it.
- Security instincts that hold up inside someone else’s production environment, including auth, secrets, and data boundaries.
- Daily proficiency with AI development tools. We weigh the quality of your direction, not whether you typed every line.
Preferred Qualifications
- Client-facing or consulting experience in any form, including internal stakeholders who behaved like clients.
- A function we sell into: customer operations, finance, revenue operations, or healthcare operations.
- MCP, agent interop protocols, or production observability and eval tooling such as tracing and trajectory analysis.
- You have mentored an engineer into being measurably better than they were.
Position Details
- Type: Full-Time
- Location: Remote (US) or nearshore
- Base Salary Range (US): $115,000 - $140,000
- Nearshore: $4,500 - $6,000 USD / month
- Reports to the Head of Engineering
Benefits
At Metacto, we believe that great work starts with a great workplace. We offer a competitive total rewards package that supports your well-being, growth, and financial security.
Our benefits include:
- 100% remote work with flexibility to manage your schedule
- Unlimited paid vacation to recharge and maintain work-life balance
- 401(k) plan with a 400% company match on the first 6% deferred
- Comprehensive medical, dental, and vision insurance
- Health Savings Account (HSA) and Flexible Spending Account (FSA) options
- Group term life insurance, plus additional coverage options
How We Hire
- A 30-minute intro with our Head of Engineering
- A look at one of your public repos, discussed with you
- A working session on a real engagement problem, sanitized
- A values conversation with our CEO
No take-home that eats your weekend. If the working session runs long at the senior level, we pay for it.
Not sure which level you are? Apply to the one closest and tell us what you have shipped. We level in the process.