What You Can Do with AI
Real tasks. Real prompts. Copy, paste, and get results in minutes.
Land a new client before lunch
How to use this
- Run PROMPT 1 in Perplexity (perplexity.ai) to research your target prospect and their company. Copy the full output.
- Open ChatGPT (chatgpt.com). Paste PROMPT 2 along with the Perplexity research. This generates your personalized outreach assets.
- Pick your best email + LinkedIn note. Personalize any details that feel generic.
- Send the email from your actual inbox. Send the LinkedIn note as a connection request. The value-first approach means no awkward "Can I pick your brain?" energy.
This is a 2-step workflow. Run each prompt in order. --- PROMPT 1: Prospect Research (run in Perplexity at perplexity.ai) --- Research [PROSPECT NAME] at [COMPANY NAME]. I need: 1. What the company does (one sentence) 2. Recent news, launches, or funding in the last 6 months 3. The prospect's role and likely priorities 4. 1–2 specific challenges or opportunities the company is facing right now 5. Their tech stack, tools, or platforms (if publicly visible) Cite sources so I can verify. --- PROMPT 2: Outreach Assets (run in ChatGPT — paste Perplexity research below) --- Using the prospect research below, write personalized outreach assets. PROSPECT RESEARCH: [PASTE THE FULL OUTPUT FROM PROMPT 1] MY OFFER: [WHAT YOU DO IN ONE SENTENCE — e.g. "I help D2C brands cut mobile load times by 40%, which typically increases conversions 15–20%"] GENERATE: A) 2 cold emails (under 120 words each) - Each must reference 1 specific finding from the research - Use the prospect's language, not yours - One CTA only: 15-minute fit check - Give 3 subject line options per email B) 2 LinkedIn connection notes (under 280 characters) - Reference something specific about them or their company - No selling — just genuine relevance C) 1 value-first email (under 100 words) - Lead with a free mini-audit or specific insight about their business - No ask for a meeting — just deliver value - End with "no reply needed" energy Rules: - Never use "I'd love to" or "I wanted to reach out" - Never use "synergy," "leverage," or "touch base" - Sound like a helpful peer, not a salesperson
Why this works: Two prompts, one complete outreach campaign. Perplexity does the research, ChatGPT turns it into personalized emails, LinkedIn notes, and a value-first audit — each referencing real data about the prospect. This replaces hours of manual research and produces outreach that actually gets replies.
Write proposals that close deals
How to use this
- After your discovery call, jot down raw notes (even messy ones work). Paste them into PROMPT 1 in ChatGPT or Claude.
- Review the structured Proposal Brief output. Fix any details AI got wrong — it's working from your notes, so check the facts.
- In the same chat, run PROMPT 2 with the brief details filled in. This generates the executive summary your buyer will read first.
- Paste the executive summary at the top of your proposal doc (Google Docs, Notion, or your proposal tool). Add pricing, timeline, and scope below it.
--- PROMPT 1: Discovery Notes to Proposal Brief --- You are a top-performing enterprise sales rep. Convert my raw discovery notes into a structured Proposal Brief. MY DISCOVERY NOTES (paste your messy notes here — bullet points, call recordings, or even stream-of-consciousness is fine): [PASTE YOUR RAW DISCOVERY NOTES — e.g.: "Talked to Sarah, VP Marketing at Bloom Health. 50-person team. They're spending $40K/month on paid ads but can't track which campaigns drive actual demos (not just clicks). Current tool is HubSpot but their attribution is broken. She mentioned they need to report ROI to the board in Q2. Budget approved for a new tool, $2K–$5K/month range. Main concern: implementation timeline — they got burned by a 6-month Salesforce rollout last year."] Output exactly in this format: 1) Buyer context (3 bullets — who they are, their situation) 2) Pains (3 bullets — in the BUYER'S exact words from discovery) 3) Success criteria (3 measurable bullets — what "win" looks like for them) 4) Proposed approach (5 bullets — how you'd solve it) 5) Options (2–3 tiers: Good / Better / Best with rough pricing) 6) Risks + mitigations (3 bullets — what could go wrong and how you'd prevent it) 7) CTA + next step (2 lines — specific, easy, time-bound) --- PROMPT 2: Executive Summary (P-A-S Framework) --- You are a sales proposal expert. Create an executive summary using the Pain-Agitate-Solution framework. INPUTS (fill in from your Proposal Brief above): Client: [COMPANY NAME] Industry: [THEIR INDUSTRY] Their Challenge: [SPECIFIC PROBLEM FROM DISCOVERY] Business Impact: [HOW THIS COSTS THEM MONEY OR TIME — use real numbers] Our Solution: [WHAT YOU DO, briefly] Proof: [YOUR BEST CASE STUDY OR METRIC] Next Step: [THE SPECIFIC ASK — e.g. "30-minute implementation walkthrough on Thursday"] FORMAT: - PAIN (2 sentences): Their quantified problem - AGITATE (3–4 sentences): What happens if they don't fix it — consequences, missed targets, board pressure - SOLUTION (2–3 sentences): How you specifically address this - EVIDENCE (1–2 sentences): Proof with real metrics from a similar client - NEXT STEPS (1–2 sentences): One specific, easy action with a date Tone: Confident, concise, no hype. Length: 250–300 words maximum. Use "you/your" language throughout — this is about THEM, not you.
Unlock this prompt
No spam. Just the prompt.
Why this works: Two prompts, two outputs. The first structures your messy discovery notes into a tight brief (even stream-of-consciousness notes work). The second turns that brief into a Pain-Agitate-Solution executive summary — the page every buyer reads first. Freelancers report significantly higher close rates with structured proposals.
Create executive decks from live data
How to use this
- Gather your weekly metrics (revenue, pipeline, risks, project status). You can pull these from your dashboard, Notion, or spreadsheet.
- Go to gamma.app. Click "Create new" > "Generate" > "Presentation."
- Copy the prompt below. Fill in the DATA INPUT section with your actual numbers. Paste into Gamma and generate.
- Review the deck, edit any slides, and export as PDF or share the Gamma link with your executive team before Monday's meeting.
Role: Chief of Staff Goal: Create a Weekly Business Review deck for the General Manager Tone: Professional, concise, data-driven. No fluff. DATA INPUT (fill in with this week's numbers): - Revenue this week: [$ AMOUNT] vs. target [$ TARGET] — [UP/DOWN X%] - Key revenue driver: [WHAT MOVED THE NEEDLE — e.g. "Enterprise deal closed: $45K ARR"] - Pipeline status: [DEALS IN PIPELINE — e.g. "12 deals, $180K weighted value"] - Top risk: [BIGGEST CONCERN — e.g. "Churn risk: 3 accounts flagged, $22K ARR at risk"] - Key project update: [MILESTONE STATUS — e.g. "V2 launch on track for March 15"] - Decisions needed: [ITEMS REQUIRING SIGN-OFF — e.g. "Approve Q2 headcount plan, sign vendor contract"] - Full metrics: [PASTE A TABLE OR LIST OF ALL KEY METRICS] SLIDE STRUCTURE (7 slides): 1. Title Slide: [THIS WEEK'S DATE] + "Weekly Business Review" 2. Executive Summary: 3 bullet points summarizing vital signs (green/yellow/red) 3. Deep Dive — Revenue: focus on the key driver, include week-over-week trend 4. Deep Dive — Risks: visualize the main concern with impact and mitigation 5. Operational Update: Key project or milestone status with timeline 6. Decisions Required: Items needing sign-off, each with context and recommendation 7. Appendix: Full metric table for reference DESIGN: - Professional dark theme - Use 2-column layouts for comparison slides - Include simple charts for revenue trends and pipeline
Available when AI School launches
This prompt will be available in the app. Join the waitlist for early access.
Join the WaitlistWhy this works: A polished 7-slide business review from your raw weekly data. Gamma handles the design. You deliver executive-ready decks every Monday without touching PowerPoint. Fill in the numbers, generate, review, send — 15 minutes total.
Never miss an action item again
How to use this
- Record your meeting with Otter.ai (otter.ai — works on Zoom, Google Meet, Teams, or in-person). After the meeting, open the transcript in Otter.
- In Otter, click the transcript text and copy it all (Cmd+A, Cmd+C on Mac / Ctrl+A, Ctrl+C on Windows).
- Open ChatGPT (chatgpt.com). Paste PROMPT 1, then paste the transcript below it. AI extracts every action item — even implied ones.
- Copy the action item table into your project management tool (Notion, Asana, Trello, etc.). At the end of the week, run PROMPT 2 with all your week's transcripts for a team digest.
--- PROMPT 1: Action Item Extraction --- Review the meeting transcript below and extract every action item — both explicit and implied. Return as a table with columns: Owner | Task | Due Date | Priority (P1/P2/P3) Rules: - If someone said "I'll do X" — that's explicit. Include it. - If someone said "We should look into Y" or "That needs to happen" — that's implied. Include it with owner marked as "TBD — needs clarification." - If no due date was mentioned, mark as "TBD" - Use YYYY-MM-DD format for all dates - Default to P2 unless urgency was clearly stated - After the table, list any decisions that were made (not action items, but conclusions reached) MEETING TRANSCRIPT: [PASTE YOUR FULL OTTER.AI TRANSCRIPT HERE] --- PROMPT 2: Weekly Meeting Digest (run at end of week) --- Review the meeting transcripts from this week (pasted below) and create a digest my team can skim in 2 minutes. OUTPUT: 1. Top 5 decisions made this week (with which meeting they came from) 2. All open action items (grouped by owner, with due dates) 3. Recurring themes or concerns (patterns across multiple meetings) 4. Meetings that still need follow-up (and what specifically needs to happen) Format: Clean, scannable, ready to paste into Slack or email. THIS WEEK'S TRANSCRIPTS: [PASTE ALL TRANSCRIPTS, SEPARATED BY "--- Meeting: [NAME] ---"]
Available when AI School launches
This prompt will be available in the app. Join the waitlist for early access.
Join the WaitlistWhy this works: Otter records and transcribes. AI extracts every action item — even the implied ones your handwritten notes would miss. The weekly digest turns 5+ hours of meetings into a 2-minute team summary. Copy the output straight into your project management tool.
Automate your most annoying recurring task
How to use this
- Go to zapier.com and create a new Zap. Set the Trigger to "Gmail: New Email" (or your email provider).
- Add an Action: "ChatGPT: Conversation." Paste the prompt below into the "User Message" field. Map {{email_body}} to the Gmail email body field.
- Add a second Action: "Notion: Create Database Item" (or Asana/Trello/Todoist). Map each output field (task_title, summary, etc.) to your task database columns.
- Turn the Zap on. Every new email is now automatically parsed into a structured task with title, summary, next action, due date, and priority — no manual work needed.
You convert messy email messages into structured task data.
THE EMAIL:
{{email_body}}
RETURN THESE FIELDS (as valid JSON):
{
"task_title": "(max 80 chars, action-oriented — start with a verb)",
"summary": "(2–3 bullet points capturing the key information)",
"next_action": "(one clear, specific action — start with a verb, e.g. 'Reply to Sarah with the revised timeline')",
"due_date": "(YYYY-MM-DD format, or empty string if no deadline mentioned)",
"priority": "(P1 = urgent + important, P2 = important not urgent, P3 = low priority)",
"category": "(one of: client_work, admin, follow_up, billing, other)"
}
RULES:
- If no deadline is mentioned, leave due_date as ""
- Default to P2 unless urgency is explicitly stated (words like "ASAP," "urgent," "by tomorrow")
- next_action must always start with a verb
- If the email is spam, newsletters, or automated notifications, return: {"task_title": "SKIP", "summary": "Automated/non-actionable email", "next_action": "", "due_date": "", "priority": "P3", "category": "other"}Available when AI School launches
This prompt will be available in the app. Join the waitlist for early access.
Join the WaitlistWhy this works: This prompt runs inside a Zapier Zap: Email arrives > ChatGPT parses it > Notion/Asana/Trello creates the task automatically. Every email gets turned into structured task data — title, priority, next action, due date — without you lifting a finger. The JSON output ensures clean integration with any task management tool.
Featured Courses
Start your freelancer plan and master these first.
Get notified when AI School launches.
Be the first to access 66+ courses on 50+ AI tools.
Join the WaitlistFrequently Asked Questions
How can freelancers use AI to find clients?
Use Perplexity for prospect research, then feed the research into ChatGPT to generate cold emails, LinkedIn connection notes, and voicemail scripts — each referencing specific signals from your research. The whole process takes about 30 minutes and produces personalized outreach at scale.
Can AI write proposals that win deals?
Yes. AI can convert discovery notes into a structured Proposal Brief, then generate a full executive summary using the Pain-Agitate-Solution framework — with quantified problems, proof with metrics, and specific next steps. Freelancers report significantly higher close rates with AI-assisted proposals.
What AI tools do freelancers use most?
ChatGPT and Claude for copywriting and proposals, Perplexity for research, Gamma for executive decks, Otter.ai for meeting notes and action items, and Zapier for automating repetitive tasks like email triage.
Can AI replace meeting notes?
Otter.ai records and transcribes meetings automatically. Then you can feed the transcript to ChatGPT to extract every action item with owner, task, and due date — even implied commitments that weren't explicitly stated.
How much time can AI save freelancers?
AI saves an average of 5.4 hours per week in knowledge work tasks (Gartner 2024). Freelancers who adopt early earn 40% more because they can take on more clients while spending less time on administrative work.










