The Challenge
Rachel, an executive coach serving CEOs at PE/VC firms, ran her practice on manual, time-intensive processes — hand-summarizing sessions, coordinating four subcontracted coaches in quarterly meetings, and sending session notes clients rarely acted on. Clients were losing value between sessions, forgetting commitments and missing opportunities for self-awareness. Rachel wanted to differentiate in a crowded market without hiring engineers or building a custom platform, but had no clear path forward.
What They Built
Mollie built a no-code automation pipeline using Granola, Zapier, Relay.app, and custom prompts to automatically flow coaching sessions from recording through transcription to personalized post-session emails and follow-up reminders, plus an AI assistant trained on the coach's methodology for between-session client support.
Mollie began with a deliberate pre-automation step: testing AI prompts manually against real coaching session transcripts before building any automated pipeline. This approach validated that outputs were genuinely useful and accurate before embedding them into workflows — preventing the common problem of automating poor-quality outputs at scale.
Three tools were prioritized based on where Rachel's time was being lost and where client value was slipping between sessions. First, a Coaching Feedback Analyzer that scored each session against ICF competencies, giving Rachel structured quality feedback without manual review. Second, an Automated Client Session Summaries pipeline: Granola captured session recordings, Zapier triggered transcript processing, and Relay.app orchestrated the generation of personalized post-session email drafts and one-week follow-up reminders that arrived in Rachel's inbox for approval before delivery. Third, "Rachel Bot" — an AI assistant trained on Rachel's methodology and approach — provided clients with between-session support, helping them recall commitments and maintain self-awareness. Human review was preserved throughout; no output was delivered to clients without Rachel's approval, ensuring quality and maintaining the therapeutic relationship.
AI Role
Rachel applied feedback insights in her very next session after the Coaching Feedback Analyzer revealed she was "giving too much advice." She also developed her own coaching rubric — something she had needed for years but never had the data to create.
Infrastructure
• Granola (session recording and transcription) • Zapier (trigger and workflow automation layer) • Relay.app (multi-step automation and output orchestration) • Google Drive (document and transcript storage) • Lovable (interface or front-end tooling)
Integration Points
• Granola recording pipeline connected to Zapier trigger on session completion • Zapier routing transcripts to Relay.app for prompt-based processing • Relay.app generating personalized email drafts and follow-up reminders into coach's inbox • Rachel Bot trained on methodology documents stored in Google Drive • All client-facing outputs routed through human review step before delivery
Non-technical founders or senior operators in service businesses (coaching, consulting, agencies) with 1–20 employees who want to implement AI quickly without hiring engineers or managing technical work.
