GTM tooling examples pop up

Two very different sales demos showed up this week, and both are trying to answer the same question: which parts of selling can be handed to software without breaking the process. One example was a thread of autonomous go-to-market workflows built around Anthropic’s Claude Code, and the other was a polished artificial intelligence sales persona campaign built around Jordan Belfort. (extruct.ai) (artisan.co) The first camp is selling automation as invisible labor. In these setups, Claude Code is used to build lists, enrich records, draft outreach, update customer relationship management systems, and push work into sequencers without a person copying data between tabs. (extruct.ai) (databar.ai) That matters because a typical business-to-business outbound workflow already touches five to eight tools, including lead sourcing, enrichment, email verification, customer relationship management software, and sequencing. Every extra handoff is a chance for records to go stale or disappear, so the pitch is not “better writing” but “fewer broken steps.” (databar.ai) The tooling underneath this is getting more concrete. Model Context Protocol, an open standard for connecting artificial intelligence agents to outside tools, is now being packaged for go-to-market teams so Claude can query a customer relationship management system, call enrichment providers, and write back results from one prompt. (databar.ai) A second layer is reusable playbooks. Extruct describes “skills” for Claude Code as plain-English files that teach the agent how to handle recurring jobs like list building, hypothesis generation, email drafting, and sending, which turns one-off prompting into a repeatable operating process. (extruct.ai) Open-source projects are now packaging that idea for non-engineers. The gtmagents repository on GitHub says its Claude-based agents cover sales, marketing, customer success, and revenue operations, with claims like generating 100 qualified leads in 5 minutes and saving 15 or more hours per week on go-to-market busywork. (github.com) The second camp is not hiding the software at all. Artisan’s new campaign puts Jordan Belfort in front of its product and says its artificial intelligence business development representative, Ava 2.0, can find leads, write emails, handle replies, and book meetings “fully autonomously.” (artisan.co) (youtube.com) Artisan is also attaching a human brand to a software claim. Belfort said on April 8 that he was joining Artisan as vice president of sales, and the company’s landing page uses his name to sell a system that promises to replace a floor full of prospecting reps with one always-on agent. (youtube.com) (artisan.co) Another startup, SLAI, is pushing the same idea even further toward a synthetic closer. Its waitlist page for “Gem” says the agent handles prospecting, qualification, objections, closing, follow-up, supports 50-plus languages, and plugs into Salesforce, HubSpot, GoHighLevel, and custom application programming interfaces. (slai.com) So the split in the market is getting easier to see. One group is building artificial intelligence as back-office plumbing that removes spreadsheet work and tool switching, while another group is building artificial intelligence as a front-stage sales character with a face, a voice, and a script people can picture replacing a rep. (databar.ai) (slai.com) (artisan.co) Both approaches are really selling the same promise: fewer humans doing repetitive prospecting. The difference is whether buyers trust an agent more when it looks like a workflow engine inside their stack or when it looks like a superstar seller who never sleeps. (github.com) (artisan.co)