Charles W.

I build AI agents and automations that run real business operations: CRM sync and migration, lead routing, document and presentation automation, and internal tools that keep working at 2am. Top Rated Plus | 100% Job Success | $100K+ earned | 1,700+ hours on Upwork. Most AI automation freelancers learned the tools last year. I've been a professional programmer since 2012; I shipped commercial video games as a lead programmer before moving into automation. I've spent the last five years building the integration layer: APIs, webhooks, databases, and workflows across platforms like Make (certified partner since the Integromat days), Zapier, n8n, Knack, and Airtable. That's why my AI builds hold up; an AI agent is only as reliable as the systems it's wired into. I also know when NOT to use AI. I scope AI to what your business actually needs: fewer failure points, lower token costs, and systems that keep running. WHAT I BUILD AI agents and agentic workflows in n8n and Make: inbox triage, lead routing, and processes that run across many steps with human approval checkpoints. AI chatbots grounded in your data (RAG) on OpenAI (GPT) and Anthropic (Claude). Workflow automation and API integration: Make, Zapier, n8n, plus HubSpot, Airtable, Knack, ConnectWise, Monday, Google Workspace, and anything with an API. Custom business apps and databases on low code platforms: Knack, Airtable, WeWeb, Bubble. Automation rescue: auditing and fixing brittle setups, then migrating legacy workflows onto an architecture built for AI. Recent work includes a CRM sync between HubSpot and Airtable, a CRM migration from Zoho to Knack, a seven page web application built in WeWeb and Airtable, and a presentation automation system I've supported across 280+ hours for a single client. I'm an automations engineer with Automaton, a Top Rated Plus agency (100% Job Success, $400K+ earned since 2019), so projects that need more hands can scale without changing teams. Every engagement starts with a process audit: I map your workflow, identify what AI can safely own, what should be handled by traditional automations instead, and where a human should stay in the loop, and quote a fixed scope. Most clients have their first automation live within a week.