Hello Remote
Remote now ships AI Agents and Remote MCP, and this role builds GTM products with Claude Code. Here's how I'd approach the stack, and the proof I've built systems like it before.
James Praise · Marketing Operations and GTM Systems
Remote shipped AI Agents, a Custom Report Builder, Remote MCP and a command-line tool in a single quarter, and this role is asked to build GTM products with AI tools rather than only administer vendor software. An agent is only as trustworthy as the data it reads. A stack that spans Salesforce, HubSpot, LeanData, Outreach, ZoomInfo, Clay and Qualified was most likely added to one tool at a time, and Remote's whole promise to customers is compliance it owns end to end. Its own lead and account data should hold up to that promise.
I'd start with a data lineage map for the one AI surface that touches the most marketing data: what it reads, where that data comes from, and what governs its consent and retention today. Next comes an audit of the stack against real usage, finding what's duplicated, what's reconciled by hand and what breaks silently. Build-vs-buy calls come after that. I'd build the connecting layer only where it costs less to maintain than another tool, and the roadmap would come from what the audit finds.
At Reform, I led the email platform migration from Userlist to Brevo. I set up email authentication (DKIM, DMARC and SPF), cleaned the list from 10,104 to 7,808 contacts and validated every workflow before cutover. Bounce rates fell 18% and email churn fell 22%. I also built Reform's affiliate program on PartnerStack from scratch.
At FunnelEnvy, I built the outbound stack on Clay, lemlist and Sales Navigator, a cross-brand measurement stack on GA4, HubSpot, Looker Studio and Lightdash, and agentic infrastructure (Claude skills plus a Webflow Model Context Protocol publishing pipeline) that turns one content asset into 10+ distribution outputs.
I also built Marketing Agent Skills, 81 Claude Code skills I use every day, 34 of them public, and the Prospect & Outbound Engine, a Python pipeline that sourced 409 companies and verified 271 emails in one run. Walkthroughs are in my AI Workflow Vault.
If this is the right problem for you, I'd like to talk about it directly.