Canibuild's sales team used to record personalised prospect videos by hand. Each one took 15 to 20 minutes, so reps did them in bursts, when they had spare time. Mark Deacon replaced that routine with a system that scripts, produces and tracks the videos, and he opens his screen to show every step. Mark is Chief AI Officer at Canibuild. He has spent his career leading sales, marketing and customer success teams, and he does not write code. He built the Marketing Hub, an internal platform where Canibuild's teams serve themselves instead of queuing for engineering. The outreach video tool is the centrepiece. A rep enters a prospect's website, or imports a list from a CSV or a HubSpot segment. The system scrapes the site, writes a script and produces a video of about a minute. It opens on the prospect's own website, then moves into the Canibuild product and ends on a call to action. Each video features an AI avatar of the account's real rep, with a HeyGen face and an ElevenLabs voice. Mark said reps supplied short recordings to set these up, and a personalised landing page goes out with the video. Mark was direct about the results. By his account, conversion is roughly on par with videos reps record themselves. The gain is volume: a handful a month before, and potentially a thousand now, at about $4 a video. Video watches trigger Slack alerts and HubSpot tasks; demo bookings create deals, so reps do not need to live in the Marketing Hub. The build-versus-buy split is useful. Mark said Canibuild's tools split roughly 50/50. It buys anything customer-facing that needs near-perfect uptime, such as its 11x AI SDR. It builds internal tools where a day of downtime is an inconvenience, not a crisis. He also showed an early version of a newsletter builder. It reads the engineering team's Jira roadmap and industry news, scores what matters for a chosen segment, and publishes through HubSpot. Mark noted that it still needs a person to pick articles and images. Automatic segment-by-segment sending is the next step. The second half widens out. Every Canibuild tool starts as a flow tested in the terminal, then gets an interface so the team can run it without waiting on Mark. Staff build dashboards and apps in VS Code using shared skills. Those skills apply templates, scan for leaked API keys, push to a central GitHub and register each tool on a master dashboard, so nothing disappears when its builder leaves. Mark described his role as enablement. He wants the people with domain expertise to build, inside guardrails, without turning his team into a new engineering queue. His advice for beginners: start with Claude Code or Codex in the terminal, tie the first build to a real business problem, and keep humans on the first 20 per cent (thinking) and the last 20 per cent (review) of the work. About Mark Deacon Mark Deacon is Chief AI Officer at Canibuild. He is a revenue leader whose background spans sales, marketing, customer success and go-to-market strategy. He runs a growing set of internal AI tools for Canibuild's teams and posts about what he is building on LinkedIn. Chapters 00:00 Cold open and intro 01:50 Why manual Loom videos did not scale 04:20 Terminal first, then a self-serve interface 05:18 From prospect website to scripted video 07:06 AI avatars, cloned voices and landing pages 12:15 Conversion parity, volume and cost per video 14:35 Watches become Slack alerts, HubSpot tasks and deals 17:43 Build versus buy: the AI SDR and enrichment 20:10 A revenue leader building without engineers 22:45 Newsletter builder fed by the product roadmap 27:22 Guardrails so everyone can build 32:22 Chief AI Officer: enablement, not a new bottleneck 38:58 Rapid fire 43:35 How to start, and the 20/60/20 rule Links and resources Mark Deacon on LinkedIn: https://www.linkedin.com/in/markpdeacon/ Canibuild: https://canibuild.com/ HubSpot: https://www.hubspot.com HeyGen: https://www.heygen.com ElevenLabs: https://elevenlabs.io 11x: https://www.11x.ai FullEnrich: https://fullenrich.com Astro: https://astro.build About the host Zachary King is the founder of Works, where he installs AI into how Australian companies operate. He has worked at the commercial pointy end of AI for almost twenty years and runs his own company largely from the command line. About the show Operator AI is the show where real operators open their screens and show exactly how they run the real parts of their business with AI. The real stack, not the hype. If you want AI installed into how your business actually operates, learn more at https://workshq.com.au/ or book a time with Zachary King at https://savvycal.com/Zachary-King/zac-at-works-30. Operator AI is part of Day One Network, the podcast and media network from W2D1 Media: https://dayone.fm/.
