AI Automation Engineer

Senior AI Automation Engineer Needed to Build a Claude-Powered Marketing Operations Agent (MCP, CRM, APIs, Multi-Agent Workflow) Experience Level: Expert Project Length: 8–12 Weeks Project Overview We're looking for an experienced AI Automation Engineer to design and build an intelligent Marketing Operations Agent powered by Claude that can automate large portions of our marketing team's daily operations. This is not a simple chatbot project. We want an AI system capable of reasoning across multiple business tools, retrieving context from our CRM and internal documentation, coordinating marketing workflows, and assisting our team with campaign planning, lead management, reporting, and customer communications. The solution should leverage Claude as the primary reasoning engine while integrating with our existing software stack through APIs and the Model Context Protocol (MCP) where appropriate. The ideal candidate has experience building production-grade AI agents, workflow automations, API integrations, CRM systems, and multi-step business processes. About Us We're a B2B SaaS company that serves marketing agencies across North America. Our marketing team currently consists of: 3 Growth Marketers 2 SDRs 1 Marketing Manager 1 Customer Success Manager We use several disconnected systems, and many of our daily tasks are still manual. We're looking to build an AI-powered operations layer that helps our team work faster while reducing repetitive work. Current Tech Stack Claude Anthropic API HubSpot CRM Slack Gmail Google Calendar Google Drive Notion Airtable Google Sheets Zapier n8n GitHub Linear Webflow Project Goals Build a Claude-powered Marketing Operations Agent capable of: Lead Intelligence Retrieve new leads from HubSpot Research company information Summarize prospect websites Identify ICP fit Assign lead scores Recommend follow-up actions Campaign Assistant Allow marketing managers to ask questions such as: Which campaigns generated the highest SQL conversion rate this month? Which landing pages are underperforming? Which leads haven't been contacted in over 14 days? The AI should gather data across multiple systems before producing answers. Content Operations The system should assist with: Writing campaign briefs Generating email sequences Creating LinkedIn posts Repurposing webinars into blogs Producing ad copy Suggesting A/B test ideas Content should follow our internal brand guidelines stored in Notion. CRM Automation Automatically: Update HubSpot properties Create follow-up tasks Assign owners Generate meeting summaries Log customer interactions Detect stale opportunities Recommend next actions Meeting Intelligence After sales calls, the AI should: Summarize meetings Extract action items Update HubSpot Notify Slack channels Schedule follow-up reminders Draft personalized follow-up emails Marketing Reporting Generate weekly reports that include: Campaign performance Lead sources Pipeline movement Attribution insights Conversion rates Top-performing channels Executive summaries Reports should be delivered automatically every Monday. Internal Knowledge Assistant Claude should answer employee questions using information from: Notion Internal SOPs Marketing playbooks Product documentation Brand guidelines Previous campaign reports Workflow Orchestration The solution should coordinate workflows across multiple systems, including: HubSpot Slack Gmail Google Drive Google Calendar Airtable Google Sheets GitHub Linear Technical Requirements The ideal solution should include: Claude as the primary reasoning model MCP integrations where appropriate Secure API integrations Modular architecture Reusable workflow components Error handling and retry logic Logging and monitoring Prompt versioning Configuration management Environment variable support Preferred Skills We're looking for someone experienced with most of the following: Claude API Model Context Protocol (MCP) AI agent architecture n8n Zapier HubSpot API REST APIs OAuth authentication Python or TypeScript Vector databases RAG architectures Prompt engineering Workflow automation Git/GitHub Docker (preferred) Deliverables The completed project should include: Fully functional AI Marketing Operations Agent Source code with documentation Workflow diagrams API documentation Prompt library Deployment guide Administrator guide User documentation Loom walkthrough demonstrating the system Test cases covering key workflows Success Criteria The solution will be considered successful if it can: Reduce manual marketing operations by at least 60% Automatically update CRM records with minimal human intervention Generate accurate campaign reports Retrieve knowledge from internal documentation with high relevance Execute multi-step workflows reliably Handle failures gracefully and provide actionable logs Be easy for our team to maintain and extend Proposal Questions When applying, please include: Describe a similar AI agent or automation system you've built. How would you structure this solution using Claude and MCP? Which parts would you implement as autonomous agents versus deterministic workflows? How would you ensure reliability and observability for production use? Which tools and frameworks would you use, and why? Please share examples of previous AI automation, CRM integration, or workflow projects.