Jainam Parmar Profile picture
Mar 16 13 tweets 5 min read Read on X
After 6 months of using NotebookLM, I can say it's the research tool that has revolutionized my workflow the most.

But only because I learned these 10 prompts.

Here's the complete system that turns 200 pages into clear answers in under an hour: Image
1. The Source Onboarding Prompt

Before you do anything else, run this the moment you upload your documents.

Paste this into NotebookLM:

"You now have access to [X] sources I've uploaded. Before I start asking questions, give me: 1) The 3 most important overarching themes that run across all these documents, 2) Where these sources agree with each other and where they contradict, 3) The single most surprising or counterintuitive finding across all of them, 4) What major questions these documents raise but don't fully answer."

This gives you a complete map of your research before you've asked a single real question.

Most people skip this. Don't.
2. The Contradiction Hunter

This is where NotebookLM becomes genuinely dangerous in the best way.

Use this:

"Go through all my uploaded sources and find every place where two or more sources disagree, contradict each other, or come to different conclusions on the same topic. For each contradiction: 1) Quote the specific conflicting claims, 2) Identify which source each claim comes from, 3) Give me your assessment of which position has stronger supporting evidence, 4) Flag this as something I need to investigate further."

The best research lives in the gaps between sources.

This prompt finds every single gap.
3. The Expert Briefing Builder

You need to explain this research to someone else — a client, a manager, a professor, an investor.

Use this:

"Based on all my uploaded sources, create a professional briefing document on [topic]. Structure it as: 1) Executive summary in 5 sentences max, 2) Key findings ranked by importance, 3) The strongest evidence supporting each finding with the source cited, 4) What remains uncertain or contested in this space, 5) 3 clear recommendations or conclusions someone should take away from this research."

Go from raw documents to a polished briefing in under 5 minutes.

This used to take hours.
4. The Question Generator

Most researchers don't know what they don't know.

This prompt fixes that.

"Based on everything in my uploaded sources, generate: 1) The 10 most important questions someone deeply studying this topic should be able to answer, 2) The 5 questions that my current sources don't fully answer but need to, 3) The 3 questions that would completely change my understanding of this topic if answered differently, 4) What a skeptic or critic would ask to challenge the main conclusions in these documents."

Use this to find exactly what's missing from your research before you go any further.
5. The Evidence Ranker

Not all evidence is equal.

Top researchers know how to tell the difference between strong evidence and weak evidence.

Most people can't.

Use this:

"Look at the key claims made across all my sources. For the 5 most important claims: 1) Tell me how strongly each claim is supported by evidence, 2) Identify whether the evidence is from primary research, expert opinion, or anecdotal, 3) Flag any claims that sound confident but are actually poorly supported, 4) Tell me which claims I should rely on heavily vs treat with caution."

Build your research on strong foundations only.

This prompt shows you exactly what's solid and what's shaky.
6. The Timeline Reconstructor

For any topic with history, chronology matters enormously.

Context gets lost when you skip the timeline.

Use this:

"Based on my uploaded sources, reconstruct the complete timeline of [topic or event]. Include: 1) The key moments, decisions, or developments in chronological order, 2) What caused each major shift or turning point, 3) How the thinking or consensus around this topic has evolved over time, 4) What the current state looks like compared to where things started, 5) What trajectory this appears to be on going forward."

Suddenly everything makes more sense when you see how it unfolded.
7. The Counterargument Shield

Before you publish, present, or defend your research run this.

"Based on my uploaded sources and the conclusions I'm drawing, help me prepare for pushback. Generate: 1) The 5 strongest counterarguments someone could make against my main conclusions, 2) The weaknesses in my evidence that a critic would immediately attack, 3) What assumptions I'm making that aren't fully proven by my sources, 4) How I should respond to each counterargument using evidence from my own documents."

Walk into every presentation, meeting, or defense already knowing every objection.

And already having the answer.
8. The Knowledge Gap Map

This prompt tells you exactly what research you still need to do.

Don't skip it.

"After reviewing all my uploaded sources, identify the gaps in my current research. Specifically: 1) What important subtopics are barely covered or completely missing from my sources, 2) What types of sources am I missing data, case studies, expert opinions, primary research, 3) Which of my conclusions feel thin because I don't have enough supporting evidence, 4) If I had to add 5 more sources to make this research bulletproof, what kind of sources should they be?"

Most researchers find out they have gaps when someone else points them out.

This prompt lets you find them first.
9. The Insight Extractor

Summaries tell you what's in the document.

This prompt tells you what it actually means.

"Go beyond summarizing my uploaded sources. I want you to: 1) Identify the 3 non-obvious insights buried in these documents that most readers would miss, 2) Find patterns across sources that aren't explicitly stated but clearly exist, 3) Tell me what the authors are implying but not directly saying, 4) Identify any data points or findings that seem small but are actually significant if you think about their implications."

The difference between a good researcher and a great one is what they see that others don't.

This prompt trains that skill.
10. The Final Report Generator

You've done the research. Now turn it into something you can actually use.

"Using all my uploaded sources and our entire conversation as context, generate a complete research report on [topic]. Include: 1) A sharp title and executive summary, 2) Key findings section with evidence cited for each point, 3) Analysis section that goes beyond the facts and explains what they mean, 4) A section on limitations and what remains uncertain, 5) Conclusion with clear takeaways and next steps. Write it in [academic/professional/conversational] tone."

From raw documents to a complete research report.

In one prompt.
Here's what most people get wrong about NotebookLM.

They think the tool is the advantage.

It's not.

The prompts are the advantage.

NotebookLM is just a container. What you put in and what you ask determines everything you get out.

These 10 prompts turn it from a document reader into a full research assistant that thinks with you, not just for you.

Save this thread. Upload your next project and run prompt 1 today.
If you made it this far, you’ll love The Shift.

We break down the biggest AI breakthroughs, tools, and strategies in under 5 minutes a day.

Join 1000s staying ahead of the curve:

Subscribe →

Plus: get 3,000+ AI tools + free AI courses when you join.theshiftai.beehiiv.com/subscribe

• • •

Missing some Tweet in this thread? You can try to force a refresh
 

Keep Current with Jainam Parmar

Jainam Parmar Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @aiwithjainam

Mar 14
🚨 Claude can now think like a Goldman Sachs analyst.

Here are 13 insane Claude prompts that build full DCF models, earnings breakdowns, and sector risk reports from public filings in minutes (Save this): Image
1. The “10-K Deep Dive” Prompt

Prompt:

"Act as a senior equity research analyst. Analyze this company’s latest 10-K filing. Extract the key revenue drivers, cost structure, operating risks, and management strategy. Summarize the insights an institutional investor would care about."

Turns a 200-page filing into actionable insights.
2. The “DCF Model Builder” Prompt

Prompt:

"Using publicly available financial data for {company}, build a discounted cash flow (DCF) model. Estimate revenue growth, operating margins, free cash flow projections, and intrinsic valuation assumptions. Explain every step."

Instant valuation framework.
Read 15 tweets
Mar 12
I collected every Perplexity AI prompt shared by quant traders and finance researchers.

These turned a search tool into a Wall Street-level research assistant.

12 copy-paste prompts.

Steal them all 👇 Image
1. Market narrative extraction

Prompt:

"Analyze the last 48 hours of financial news and research reports about {company/sector}. Identify the 3 dominant narratives driving investor sentiment and explain the evidence behind each narrative."

Why quants use it:

Find the story moving the market before it shows up in price.
2. Institutional thesis breakdown

Prompt:

"Find hedge fund letters, analyst reports, and institutional commentary on {company}. Extract the strongest bull thesis and bear thesis. List the assumptions behind each."

This surfaces how professionals actually think about the trade.
Read 15 tweets
Mar 10
BREAKING: Claude can now research like a Stanford PhD student.

Here are 9 insane Claude prompts that turn 40+ research papers into structured literature reviews, knowledge maps, and research gaps in minutes (Save this) Image
PROMPT 1 - The Intake Protocol

Use this when you first upload your papers:

"I'm going to share [X] papers on [topic].
Before I ask anything, do this:

1. List every paper by author + year + core claim in one sentence
2. Group them into clusters of shared assumptions
3. Flag any paper that contradicts another

Don't summarize. Map the landscape."
PROMPT 2 - The Contradiction Finder

Most researchers miss this. This prompt doesn't:

"Across all papers uploaded, identify every point where two
or more authors directly contradict each other.

For each contradiction:
- State both positions
- Name the papers
- Explain WHY they likely disagree (methodology, dataset, era)

Format as a table."
Read 12 tweets
Mar 6
Most people write prompts telling AI what to do.

The ones getting 10x better outputs also tell it what NOT to do.

It's called negative prompting and it takes 30 seconds to learn.

Here's a step-by-step guide: 👇 Image
Step 1: Understand what negative prompting actually is.

A normal prompt says: "Write me a product description."

A negative prompt says: "Write me a product description. Don't use hype words, don't use bullet points, and don't sound like a sales ad."

You're giving the AI a guardrail, not just a goal.
Step 2: Before you write anything identify what annoys you most about AI outputs.

Common ones:

→ Too formal / too robotic
→ Adds bullet points when you wanted paragraphs
→ Repeats your question back to you
→ Uses words like "certainly!" or "absolutely!"

Write those down. They become your negative prompt.
Read 11 tweets
Mar 5
After 2 months of using Claude Cowork daily, I can say it's the tool that has changed how I work more than anything else.

So here are 10 mega prompts that automated my entire business and could do the same for you: Image
PROMPT 1: BULK CONTENT PRODUCTION SYSTEM

---


You are a world-class content production director who has
scaled content operations for 8-figure media companies. You
produce platform-native content that drives engagement, saves,
and shares — never generic filler.



I am uploading a folder containing 25 raw topic briefs as
text files. You will process EVERY file — no skipping,
no summarizing, no combining topics.



For each topic brief, produce the complete content package:

1. X THREAD (10 tweets)
— Tweet 1: Viral hook using one of these formats: shocking
stat, contrarian claim, story open, or insider reveal
— Tweets 2–9: One concrete insight per tweet, each ending
with a bridge line that forces the next read
— Tweet 10: CTA with engagement trigger ("Save this" /
"Comment X for Y")

2. LINKEDIN POST (200–250 words)
— Hook line that stops the scroll
— 3-paragraph body using the problem → insight → application
structure
— Closing line with a question to drive comments

3. INSTAGRAM REELS SCRIPT (60 seconds)
— Written in Hinglish where natural
— Hook in first 2 seconds (spoken line + visual direction)
— 5–6 punchy beats with b-roll notes
— Closing CTA with voiceover direction

4. 7 HOOK VARIATIONS
— Each under 12 words
— Use different formats: stat, question, contrarian,
story, list tease, insider, fear

5. EMAIL SUBJECT LINE (5 variations)
— Under 9 words each
— Include one curiosity gap, one urgency, one social proof



— Label every output: TOPIC [NUMBER] → [FORMAT]
— Output all 25 packages back to back in one continuous response
— Do not add commentary between topics
— Every output must be ready to copy-paste with zero editing
— Do not reduce quality on topics 10–25. Maintain identical
depth throughout.
PROMPT 2: FULL BUSINESS INTELLIGENCE AUDIT

---


You are a senior partner at a top-tier strategy firm. You
have led 200+ business audits across SaaS, e-commerce, media,
and services companies. You identify what's bleeding revenue,
what's blocking growth, and exactly what to do about it —
with specifics, not platitudes.



I am uploading:
— 12 months of revenue and expense data (CSV)
— 6 months of customer support tickets (exported text)
— All email campaign reports (PDF)
— 3 months of social analytics exports
— NPS survey responses (CSV)
— Top 5 competitor landing pages (screenshots)




1. REVENUE LEAK REPORT
— Identify every place money is leaving the business
— Quantify each leak with a dollar estimate where possible
— Rank leaks by size: Critical / Major / Minor
— Include the specific file reference that evidences each leak

2. GROWTH BLOCKER ANALYSIS
— Top 5 things preventing the business from scaling
— For each: root cause, downstream effects, fix priority

3. CUSTOMER INTELLIGENCE SUMMARY
— Segment customers into 3–5 distinct behavioral profiles
using ticket + NPS data
— For each segment: what they love, what they complain about,
what they actually want that we're not giving them
— Identify the highest-LTV segment and what makes them sticky

4. COMPETITIVE GAP MAP
— Where each competitor is stronger
— Where we have genuine advantages they're ignoring
— 3 positioning angles we can own that none of them have claimed

5. 90-DAY TURNAROUND PLAN
— Week-by-week action plan
— Each action tagged: Revenue Impact (High/Med/Low) +
Effort Required (High/Med/Low)
— First 2 weeks must be executable with existing team,
zero new hires

6. EXECUTIVE ONE-PAGER
— Summarize everything above in 400 words
— Written for a board audience: direct, no jargon,
conclusion-first



— Cite specific data points from uploaded files throughout
— Do not make assumptions without flagging them explicitly
— If data conflicts across files, surface the conflict
and explain which source to trust and why
— Output each deliverable as a clearly labeled section
Read 13 tweets
Feb 27
🚨 BREAKING: Someone just leaked their full Claude Cowork setup and it compresses an entire workday into 90 seconds.

I scraped every power user workflow across X, Reddit, and private Slack groups to find out how.

99% of people are using it completely wrong.

Here's what the top 1% actually do 👇Image
Prompt 1: Inbox triage + summarization

"You are a Chief of Staff with 10 years of executive support experience.

I need you to process my inbox one email at a time using this exact chain of reasoning:

Step 1 → Classify: Is this urgent (needs reply today), important (needs reply this week), or noise (unsubscribe/archive)?
Step 2 → Extract: Pull out the sender, request, deadline, and any names mentioned.
Step 3 → Draft: Write a reply under 4 sentences. Match the sender's tone. Never use "I hope this email finds you well."
Step 4 → Flag: If it involves money, legal language, or a deadline under 24 hours, mark it [ESCALATE] before the reply.

Process every email in my inbox folder. Output in this format:
[CLASSIFICATION] | [EXTRACTED INFO] | [DRAFT REPLY] | [FLAG IF NEEDED]

Do not stop until every email is processed."
Prompt 2: Document drafting

"You are a McKinsey Senior Consultant who writes exclusively in plain English.

Hard constraints:
- No bullet points. Prose only.
- No sentence longer than 20 words.
- Every paragraph must end with a decision or action.
- Never use: "leverage," "synergy," "circle back," or "moving forward."

Task: Read every file in my [Documents/Drafts] folder. For each document that is more than 50% complete, turn it into a final polished version ready to send.

For documents under 50% complete, write a 3-sentence brief explaining exactly what still needs to be done and who should do it.

Save all outputs as [filename]_FINAL.docx in the same folder."
Read 12 tweets

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Don't want to be a Premium member but still want to support us?

Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal

Or Donate anonymously using crypto!

Ethereum

0xfe58350B80634f60Fa6Dc149a72b4DFbc17D341E copy

Bitcoin

3ATGMxNzCUFzxpMCHL5sWSt4DVtS8UqXpi copy

Thank you for your support!

Follow Us!

:(