How I AIHow to turn meeting notes into prototypes that your sales team can immediately demo to customers
CHAPTERS
Meet Anjan (CPTO) and the core promise: faster alignment with AI prototypes
Claire Vo introduces Anjan Panneer Selvam (CPTO of Acolyte Health) and frames the episode around a new operating model: using AI to turn stakeholder conversations into interactive prototypes that sales and customers can use immediately. The emphasis is not just speed, but reducing friction across product, engineering, and go-to-market teams.
- •CPTO perspective: owning both product and engineering outcomes
- •AI embedded in stakeholder interactions (CEO, engineering, sales, customers)
- •Prototypes used as a communication and alignment artifact
- •Theme: move from documents/debates to demos/alignment
- •Tooling + culture change are both required
How AI blurs product and engineering boundaries (and why CPTOs thrive)
Anjan explains how AI reduces the “translation cost” between ideas, product intent, and engineering execution—making the combined CPTO role more practical. He argues AI shifts focus away from process overhead toward making possibilities tangible quickly.
- •AI cuts time spent transitioning ideas into structured artifacts
- •Early-stage startups benefit from fewer handoffs and faster decision loops
- •Roles blur: product thinking and technical exploration converge
- •AI makes it easier to carry dual responsibilities (product + tech)
- •Outcome focus: from ‘what should we do’ to ‘what’s possible quickly’
Stakeholder idea → prototype workflow: brain dump to structured build prompt
The workflow begins in a live stakeholder meeting (often with a CEO request like “build a journey builder”). Anjan captures rough natural-language input, uses ChatGPT to normalize it into a structured prompt, then feeds that prompt into prototyping tools like Lovable or v0 to generate an interactive experience.
- •Start with unstructured conversation; avoid forcing PRD structure too early
- •ChatGPT used to convert vague requests into a tool-ready prompt
- •AI helps “normalize” product sense into consistent, structured output
- •Prototype-first approach reduces weeks of alignment into minutes/hours
- •Designed for high-stakes meetings where misalignment is costly
Capturing meeting transcripts with the Limitless Pendant for instant AI prompting
Anjan demonstrates using the Limitless Pendant wearable to record meetings and generate transcripts with minimal friction. Those transcripts become immediate inputs to ChatGPT—turning spontaneous stakeholder conversations into prototype-building prompts.
- •Wearable recording is non-intrusive compared to laptop/phone setups
- •Fast capture of ideas in ad-hoc settings (coffee chats, hallway talks)
- •Transcript → ChatGPT prompt pipeline removes reliance on memory/notes
- •Highlights the coming shift toward AI-native hardware in workflows
- •Compliance note: disclose recording appropriately
Building an interactive workflow/journey builder prototype in Lovable (with React Flow)
With the structured prompt, Anjan builds a high-fidelity interactive prototype in Lovable—enhanced by specifying implementation details like React Flow for a canvas-based builder. He iterates progressively, starting simple and layering complexity, until the prototype supports drag-and-drop journey construction.
- •Lovable chosen for polished UI and interactivity out of the box
- •Add specific libraries (e.g., React Flow) to guide the implementation
- •Iterate by breaking the concept into smaller parts for higher success
- •Prototype becomes editable and explorable—not a static mock
- •Demonstrates feasibility and quickly surfaces missing requirements
Why prototypes beat documentation for complex logic-heavy products
Claire and Anjan argue interactive prototypes are especially powerful for complicated rule builders where documentation becomes unreadable. Clicking through behavior exposes edge cases more naturally than reading long PRDs, enabling cheaper failure and faster correction with executives and engineering.
- •Complex logic is hard to express and validate via documents alone
- •Interactive prototypes act as a more natural “requirements test”
- •Cheaper path to failure: wrong direction discovered quickly
- •Reduces 17–18 page specs, sketches, and slow handoffs
- •Aligns CEO + engineering with a concrete demo rather than abstractions
Market research in minutes: Perplexity deep research + ‘devil’s advocate’ prompting
After prototyping, Anjan validates demand and monetization potential using Perplexity for fast market/competitive research. Because AI can be overly optimistic, he explicitly asks for pushback and counterarguments to decide what to build—and what to reject—faster.
- •Use the same core prompt to generate a market research brief
- •Perplexity deep research compresses analysis time dramatically
- •Add prompts to challenge assumptions (devil’s advocate, pro/cons)
- •AI helps say “no” earlier without weeks of sunk effort
- •Goal: build conviction (or disconfirm) before investing engineering time
Turning research into stakeholder-ready decks with Gamma
Anjan converts the validated concept and research into polished slides using Gamma, making it easier to communicate strategy to execs and cross-functional partners. The deck helps signal preparedness and creates a shared narrative around value, differentiation, and next steps.
- •Prompt reuse: prototype concept → research → deck pipeline
- •Slides cover competitive landscape, strategic benefits, and product intent
- •Polish matters: faster buy-in when artifacts look complete
- •Decks reduce the burden of “selling” the idea internally
- •Supports rapid stakeholder alignment beyond product/engineering
ChatPRD as the requirements gate: validating details before engineering execution
To handle inevitable follow-up questions (“what exactly do you mean?”), Anjan uses ChatPRD to pressure-test requirements. He relies on its structured validations, aligns a repeatable PRD template with his team, and produces concise, readable specs with a clear v0/v1/v2 framing.
- •ChatPRD used as a ‘gate’ to ensure key questions are answered
- •Structured validation prompts catch gaps and risks early
- •Team-aligned PRD template (“CPTO stack”) standardizes expectations
- •Concise bullets over long paragraphs; avoids premature Jira sprawl
- •Captures “future ideas/enhancements” to guide roadmap thinking
A living demo/prototype library for sales and customer success to use with customers
Anjan describes deploying prototypes as microsites that mirror the product’s design language, creating a “living product library.” Sales and customer success can demo these reliably (unlike brittle Figma clickthroughs), gather real-time feedback, and accelerate customer alignment without engineering interruptions.
- •Company-wide accessible prototype links for sales/CS/execs
- •Interactive demos reduce navigation friction and increase confidence
- •Real-time customer feedback within minutes of stakeholder meetings
- •Avoids pulling engineers into one-off demo building
- •Cultural shift: more openness with roadmap/prototypes builds partnership
Breaking engineering deadlocks with mobile prototypes using Rork (no mobile dev needed)
When engineering resists an idea due to missing mobile expertise, Anjan uses Rork to prototype a simple mobile app (e.g., capturing selfies for avatar expressions) to prove feasibility. The intent isn’t to replace engineering, but to unblock uncertainty and reframe the conversation around what it would take to productionize.
- •Use case: mobile capture to overcome poor enterprise laptop webcams
- •Rork enables quick iOS/Expo-style prototypes via prompting
- •Prototype used to demonstrate possibility, not claim production readiness
- •Reframes ‘we can’t’ into ‘here’s what it would take’
- •Helps teams explore without premature hiring or heavy commitment
Operating model takeaways: AI elevates PMs and speeds org-wide alignment
They zoom out to the organizational implications: companies will spend more time upfront validating ideas, aligning teams, and training support/sales earlier—before production build. AI doesn’t remove the need for strong product sense; it shifts PMs toward outcomes, precision, and faster iteration cycles.
- •New bottlenecks shrink: faster validation, faster cross-functional alignment
- •Prototypes become training tools for support and enablement pre-launch
- •Small teams can move faster—alignment becomes a competitive advantage
- •AI elevates PMs rather than replacing them (focus on outcomes)
- •Goal: ship with more precision after earlier, cheaper learning
When AI outputs aren’t good: step back, switch tools, and iterate smarter
Anjan acknowledges failure cases: many prototype attempts don’t work on the first try. His strategy is to pause, change approach (often returning to ChatGPT to clarify), and avoid endless iteration inside a single tool—balancing patience with practicality.
- •Expect failures: multiple attempts may be needed for a ‘perfect’ demo
- •Take breaks to avoid unproductive prompt spirals
- •Use ChatGPT as a broader thinking layer before tool-specific iteration
- •Switch tools/workflows when stuck rather than brute-forcing
- •Net benefit still outweighs the misfires
