Agentic Engineer

Octocrew is a consulting practice by O-CMO that builds AI-native systems for B2B companies. We work at the intersection of marketing, operations, and applied AI — designing agentic workflows that replace manual processes with autonomous, composable systems.

We are looking for an Agentic Engineer — a builder who can design, ship, and maintain production-grade AI agent systems using modern orchestration frameworks and tool-use patterns.

What you will build

• Multi-agent systems using Claude API, OpenAI Agents SDK, LangGraph, CrewAI, or similar frameworks
• Tool-use pipelines: MCP servers, function calling, API integrations (Notion, HubSpot, Google Workspace, Slack, etc.)
• RAG pipelines with vector databases, chunking strategies, and retrieval evaluation
• Workflow automation that connects AI reasoning with real business actions (CRM updates, content publishing, data enrichment)
• Evaluation and observability layers: structured logging, trace analysis, quality scoring

What we expect

• Strong Python skills — clean, testable, production-ready code
• Hands-on experience building LLM-powered applications (not just wrappers around ChatGPT)
• Familiarity with at least one agent orchestration framework (LangGraph, CrewAI, OpenAI Agents SDK, Autogen, etc.)
• Understanding of prompt engineering patterns: chain-of-thought, tool-use, structured output, few-shot
• Experience with API integrations and webhook-based workflows
• Ability to work autonomously: take a brief, ask the right questions, ship working code
• English B1+ — you will read and write technical documentation in English
• Comfort working async with a small distributed team

Nice to have

• Experience with MCP (Model Context Protocol) servers
• Background in DevOps or infrastructure (Docker, CI/CD, cloud deployment)
• Familiarity with Notion API, HubSpot API, or Google Workspace APIs
• Experience with evaluation frameworks (ragas, deepeval, custom evals)
• Understanding of B2B marketing or sales operations (helps you build better agents for our clients)
• Contributions to open-source AI/ML projects
• Experience with TypeScript/Node.js for full-stack agent interfaces

Not this role if

• You are mainly a data scientist or ML researcher looking for training/fine-tuning work — we focus on applied agent engineering, not model development
• You prefer detailed specs and close supervision — we need self-directed builders
• You have no experience with LLM APIs beyond ChatGPT UI — we need hands-on API and framework experience
• You are looking for a full-time permanent position — this is a flexible contractor engagement

How to apply

Fill out a short application form: Application Form

We review applications weekly and respond to every candidate.