AI/Cloud Infrastructure Consultant

Supercell Health runs 90-day client health optimisation programs covering diagnostics, blood analysis, InBody scanning and personalised clinical protocols. We use Claude Code, Python and Google Drive to run this with a small team, and we're looking for an experienced consultant to take over the technical build from our current developer and keep it running long-term. Our current consultant has already built substantial parts of the system. This role starts with a review and handover, then moves into ongoing weekly maintenance and further development. What the role involves - Complete the clinical protocol pipeline: move from approved client data to a finished protocol and branded client document, currently built with Claude Projects, custom JSON knowledge files and Python scripts, with human review before anything reaches a client - Build and maintain a client dashboard (Replit, Vercel or similar) showing each client's current 90-day program phase, upcoming check-ins, re-test dates and protocol status - Diagnose why Claude agents or automations break down or underperform, and fix them. - Keep the code repository and its packages current, and manage GitHub pushes and syncs. - Write clear SOPs in Markdown and Google Drive for each completed system, and convert them into Claude or Cowork skills where that makes sense. - Once the handover is complete, spend a couple of hours a week on ongoing system maintenance. What you'll need - Practical experience with Claude Code, built on a solid grasp of core software engineering principles - Python, including environment setup, running scripts such as PDF generation tools, and resolving interpreter issues - Git and GitHub, with experience pushing to and syncing a shared repository - API integration experience - Experience deploying user-facing dashboards on Replit, Vercel, or a similar platform - An understanding of how to manage LLM token usage so automations don't burn through cost unnecessarily - Google Drive, Google Sheets and API integrations - Experience writing technical SOPs and developer handover documentation - Comfortable handling API keys, service accounts and other sensitive data securely How we work Each task starts with a written problem statement, an agreed output and a test process. We build, verify and document one module before moving to the next. You'll review the existing repository and workflows, work alongside the current consultant during the transition, and produce a handover clear enough that a new AI operator could pick up the work without relying on anything undocumented. Please include in your application - Examples of AI automation or infrastructure systems you've built - Your experience with Claude Code, Python, GitHub, and Google Drive - An example of how you've reduced token usage or cost in an LLM-based workflow - Your approach to taking over an existing, partially documented system - Your availability, including for the ongoing weekly maintenance component - Your hourly rate or fixed-price proposal for the initial technical review and handover This starts as a project-based engagement covering the technical review and handover, with ongoing weekly maintenance and further development once that's done.