AI Automation Engineer – Hermes, ChatGPT, Claude, MCP & ClickUp

We are looking for a hands-on AI Automation Engineer with direct experience building and operating workflows with **Hermes**. Hermes is already part of our AI stack alongside ChatGPT, Claude, MCP integrations, and ClickUp. Our goal is to make Hermes work reliably across our ClickUp projects, recurring tasks, and documented SOPs. You will improve our existing implementation, build new workflows, and ensure the system operates efficiently without unnecessarily consuming AI credits or compute resources. This is an implementation role—not a strategy or consulting-only engagement. We need someone who can understand a documented business process, build the workflow, test it, monitor it, and continue improving it after launch. ## What We’re Building ClickUp is becoming the operational control center for our company. We are documenting recurring tasks, projects, responsibilities, and SOPs so AI agents can complete or prepare as much of the work as possible. A typical workflow might: 1. Identify a recurring or assigned ClickUp task. 2. Read the task requirements and related SOP. 3. Gather data from Amazon, Meta, TikTok, or another platform. 4. Clean, transform, and organize the data. 5. Complete all or part of the required analysis. 6. Return the results to ClickUp as comments, links, reports, or attachments. 7. Update the task status and clearly identify anything requiring human review. Not every workflow must complete the entire task. Automating the data gathering, transformation, and preparation stages is still valuable if it allows a team member to focus only on analysis and decision-making. ## Primary Responsibilities - Integrate Hermes with ClickUp projects, tasks, recurring workflows, and SOPs - Enable agents to retrieve task instructions and supporting documentation - Return structured outputs, comments, links, reports, and attachments to ClickUp - Trigger workflows based on schedules, task statuses, assignments, or other events - Build AI agents from existing ClickUp tasks, SOPs, and process recordings - Connect Hermes, ChatGPT, and Claude with company tools through MCP, APIs, and webhooks - Automate recurring data collection, transformation, reporting, and task execution - Add human-review and escalation steps when a workflow cannot be completed reliably - Test workflows for accuracy, reliability, security, and appropriate error handling - Monitor, troubleshoot, document, and improve agents after deployment - Collaborate with the team members who currently own each process - Expand successful workflows across marketing and eventually the wider company ## Usage and Cost Optimization A major part of this role is ensuring our agents operate efficiently. We are currently experiencing unnecessary credit consumption and need someone who understands how to control AI usage without sacrificing output quality. You will be expected to: - Review workflows for unnecessary agent calls, repeated context, and inefficient loops - Select the appropriate model for each task based on complexity, cost, and required quality - Reduce unnecessary token and credit consumption - Improve prompts, context management, tool selection, and workflow structure - Use deterministic scripts or traditional automation when an AI call is unnecessary - Establish limits, stopping conditions, retry rules, and human-approval checkpoints - Prevent duplicate executions and uncontrolled agent behavior - Track usage and cost by workflow where possible - Measure whether each automation produces enough operational value to justify its cost - Recommend practical ways to improve reliability while reducing ongoing usage ## Required Experience Candidates must have: - Direct, hands-on experience working with Hermes - Experience connecting AI agents to external tools and operational systems - Experience with ClickUp integrations, automations, or its API - Experience building agentic workflows using ChatGPT and Claude - Working knowledge of MCP, APIs, webhooks, authentication, and tool calling - Experience managing token usage, model costs, retries, and agent execution limits - Experience automating recurring data-gathering and reporting workflows - The ability to test, monitor, troubleshoot, and maintain agents after deployment - Scripting experience with Python, JavaScript, or a comparable language Experience with ecommerce or marketing data from Amazon, Meta, or TikTok is strongly preferred. ## What Success Looks Like During the initial engagement, you will: 1. Review our current Hermes and ClickUp implementation. 2. Identify the causes of excessive credit consumption and recommend improvements. 3. Establish reliable communication between Hermes and our ClickUp tasks and SOPs. 4. Select several high-value workflows from our existing ClickUp backlog. 5. Build agents that complete or prepare those tasks reliably. 6. Deliver useful, structured results directly into ClickUp. 7. Add appropriate monitoring, error handling, and human-review steps. 8. Document how each workflow operates and how it should be maintained. 9. Establish a repeatable and cost-conscious process for deploying additional agents. ## Working Arrangement This will begin as an ongoing part-time contract, with the potential to expand significantly based on performance and the number of workflows available. You should be comfortable with: - Weekly or biweekly meetings - Asynchronous communication through Slack - Independently turning business requirements into working implementations - Collaborating with team members who can demonstrate existing processes - Maintaining and improving workflows after launch - Communicating clearly about reliability, limitations, usage, and cost ## Required Loom Video Please include a link to a 3–5 minute Loom video with your application. Applications without a Loom video—or without demonstrated Hermes experience—will not be considered. In your video, please: 1. Introduce yourself and explain why you are a strong fit. 2. Describe your specific experience with Hermes, including how recently you used it. 3. Demonstrate a Hermes implementation or workflow you personally built, if confidentiality permits. 4. Walk us through one or two recent AI-agent, automation, or data-workflow projects. 5. Explain the business problem, technical approach, and what you personally implemented. 6. Describe the measurable value delivered, such as time saved, manual work eliminated, improved accuracy, or reduced AI usage costs. 7. Explain how you tested, monitored, and maintained the workflow after launch. 8. Share an example of how you reduced token, credit, or compute usage in an AI workflow. If confidentiality prevents you from showing the original project, provide a sanitized demonstration or clearly explain its architecture, tools, and your personal contribution. Your video does not need to be highly produced. We are looking for clear communication, technical ownership, verifiable Hermes experience, and evidence that you can build dependable, cost-efficient systems. Along with your Loom video, please include: - Your availability and preferred working hours - Your hourly rate - Links to relevant portfolio materials or case studies - A list of the tools and technologies used in the projects demonstrated