When AI stalls after the pilot, the instinct is to blame the tool. CPA-turned-AI-builder Alyssa Clarcq has a blunter diagnosis: if your process isn’t documented, AI can’t follow it, just like an intern you never properly trained.
Host Jess Lampe of Launch Consulting talks with Alyssa about her path from Big Four external audit to vibe-coding an “AI-native company” with no programming background, guided largely by software developers in her TikTok comments. She explains how her system works: a second brain holding all her context, a central task tracker, specialized “desks” for finance, marketing, and content, and agent loops for prep, review, and QA. She also covers why governance and guardrails matter when AI can hallucinate on a regular basis.
They get into the realities: calculating a net savings of about 10 hours a week, building a “babysitter” chat to manage parallel agents, the surprising cognitive load of constant decision-making, and rising token costs that have her weighing local models. Alyssa closes with advice for organizations: treat AI adoption as upskilling rather than replacement, invest in real training, and start with your internal audit team, the people who already understand your processes end to end.
Podden och tillhörande omslagsbild på den här sidan tillhör
Launch Consulting. Innehållet i podden är skapat av Launch Consulting och inte av,
eller tillsammans med, Poddtoppen.
