Use NotebookLM as a consistent additional judge by loading every pitch separately, applying an anchored rubric, and generating an evidence-linked ranking that complements human judges.
Source episode · 32:00
How I AI: Marily Nika’s AI Native PM Workflow (Perplexity, Veo, v0, Notebook LM)with Claire Vo
Marily uploads each demo-day pitch as a separate NotebookLM source, asks it to rank entries against innovation, impact, and storytelling, and presents the result through an interactive audio overview.
A criterion-level scorecard and entertaining audio summary that cites the pitches, exposes uncertainty, and supports rather than replaces the accountable judging decision.
Google's AI-powered research assistant
Step by step
Follow the sequence once, then adapt the prompts, checks, and handoffs to your own setup.
3 steps
Upload one clearly named audio or transcript source per eligible pitch and reconcile the NotebookLM source list with the competition roster before any scoring.
This method works with many source types, not just audio. You could upload pitch decks, business plans, or video transcripts as well.
Give NotebookLM an anchored rubric with weights, evidence requirements, missing-evidence handling, and tie rules. Request a criterion table before requesting winners.
Evaluate every eligible pitch against this rubric: [criteria, anchors, and weights]. For each entry, cite specific source evidence, score each criterion, mark missing evidence, note uncertainty, and calculate the weighted total. Do not infer founder ability from accent, fluency, production quality, or demographic signals. Return ties where the evidence does not separate entries.
Be as specific as possible with your criteria. The more detailed your instructions, the more accurate and relevant the AI's analysis will be.
Inspect the scorecard with human judges, correct source or scoring errors, resolve ties, and then generate an audio overview that explains the selected entries and criteria. Disclose that AI supplied an additional analysis and record the human-owned final decision.
Create a concise audio overview of the final results using the approved scorecard. Explain each selected entry’s evidence against the rubric, mention any close tie or uncertainty, and state that the final awards were confirmed by the judging panel.
Using the live radio show feature can be a fun and memorable way to conclude a hackathon or demo day, making the AI judge a part of the event itself.
Turn an idea into a PRD, user stories, and a plan.
After the steps
How to recover when the loop fails and where human judgment helps.
Recover
Keep building
Log your workouts effortlessly without ever opening a spreadsheet. This workflow uses the Claude mobile app to parse a simple voice note and automatically update your workout tracking spreadsheet in the background.
Create a custom email management interface within Claude to read, triage, and draft replies without opening the Gmail tab. This workflow helps you reclaim focus and compound your learnings by training your AI on every interaction.
Build a fully automated system in Claude to generate personalized, interactive HTML proposals and client onboarding materials, saving hours on administrative work and creating a high-touch experience.
Join 100,000+ product managers who use ChatPRD to write better docs, align teams faster, and build products users love.