AI Claude Co-Work Set Up

AI Implementation Specialist — Claude Cowork & ChatGPT Work (Ongoing, Done-For-You Client Delivery) Snapshot: We're an Australian AI automation agency that builds and runs done-for-you AI systems for finance and property businesses. We're scaling client delivery and need a proven, hands-on implementer to build and ship Claude Cowork and ChatGPT Work deployments with us — from scoping a client's workflows to installing, integrating, documenting, training, and maintaining. This is ongoing contract work with room to grow as our client base does. Please read the "How to apply" section carefully — we filter hard, and generic proposals are ignored. About us We set up and manage AI systems — primarily Claude Cowork and ChatGPT Work — for Australian mortgage brokers, financial advisers, and property/commercial finance firms. Our clients are time-poor and admin-heavy, and we install AI that takes the repetitive work off their plate: file/compliance notes, client update emails, document chasing, lender/policy lookups, database reactivation, and more — all wired safely into their existing tools. You'd be the person who makes those systems actually work inside a real business's stack. What you'll be doing Set up & configure Claude Cowork and ChatGPT Work for client teams: plans, seats, workspaces, admin controls, security and permissions. Build custom agents, skills and workflows: Cowork skills & plugins; ChatGPT Work workspace agents and custom GPTs; automations that run reliably in production (not demos that break in week two). Integrate with client tools via MCP and native connectors: Google Workspace, Microsoft 365, Slack, Notion, Drive, and CRMs such as GoHighLevel, HubSpot, and Australian broking platforms (AFG, Connective, Mercury Nexus, Salestrekker, BrokerEngine). Wire up cross-system automation: APIs, webhooks, OAuth scopes, and tools like Zapier / Make / n8n where needed. Scope & map: join (or work from notes of) client discovery, identify the highest-value automations, and translate messy real-world processes into working systems. Governance & security: design least-privilege access, handle sensitive data correctly, and keep everything auditable — our clients hold sensitive financial data, so this is non-negotiable. Document & train: clear SOPs, handover docs, and short training sessions so client staff actually adopt what you build. Maintain & improve: monitor, troubleshoot, tune, and roll out new skills as connectors and models change. Must-have experience (this is the bar for "proven") A real, demonstrable track record implementing AI in actual businesses — with examples, case studies, or references we can verify. Courses and theory alone won't cut it. Deep hands-on experience with Claude Cowork: skills, plugins, MCP connectors, Projects, and Team/Enterprise admin (SSO, seats, permissions, usage). Deep hands-on experience with ChatGPT Work / ChatGPT Enterprise & Business: workspace agents, connectors, custom GPTs, enterprise admin controls (Codex experience a plus). Strong grasp of MCP (Model Context Protocol) — connecting agents to real tools and data, and building or configuring connectors/servers. Integrations & automation: confident with APIs, webhooks, OAuth, and platforms like Zapier / Make / n8n. Security & privacy discipline: least-privilege design and safe handling of sensitive/confidential data. Excellent written English, clear communication, and genuinely strong documentation habits. Self-managing and reliable — you own outcomes and communicate proactively, without hand-holding. Nice-to-have Experience in financial services, mortgage broking, lending, or property — Australian experience is a big plus. Familiarity with Australian privacy obligations (Privacy Act 1988 / Australian Privacy Principles) and broker/adviser compliance context (e.g., Best Interests Duty) — or a clear ability to work within a client's compliance rules. CRM / marketing-automation experience (GoHighLevel, HubSpot, and broking CRMs). Light coding (Python / JavaScript), RAG, prompt engineering, or self-hosting n8n. Prior agency or done-for-you delivery experience — working to briefs and deadlines across multiple clients. Tools you'll work across Claude (Cowork, Projects, MCP) · ChatGPT Work / Enterprise · Google Workspace · Microsoft 365 · Slack · Notion · Zapier / Make / n8n · GoHighLevel / HubSpot · Australian broking CRMs. This role is NOT for you if… You're a generalist "AI consultant" who has never shipped a live Cowork or ChatGPT Work build. You only write prompts or produce content and haven't done real implementation/integration work. You can't point to actual systems you've built and deployed. You need constant direction to make progress. Engagement & logistics Type: Ongoing contract (steady, growing work as our client base scales). [Set: hourly or fixed-per-project] Hours: [e.g., 5–20 hrs/week to start, scaling with demand] Location: Remote / global, but you must have some overlap with Australian business hours (AEST/AEDT) for client onboarding and team syncs. [adjust if not required] Start: ASAP How to apply (read carefully — we filter hard) Start your proposal with the word "COWORK" so we know you read this. Share 2–3 AI implementations you've personally shipped — ideally at least one Claude Cowork build and one ChatGPT Work / Enterprise build. For each, tell us: the client/context, what you built, the tools you connected, and the outcome it produced. Include a short 3–5 minute Loom walking us through one of those builds — show the actual setup, not slides. Answer briefly: How would you connect Claude Cowork to a client's Gmail and CRM with least-privilege access, and build a skill that drafts compliant client-update emails? How do you keep sensitive client data safe when deploying cloud agents?