Prompts By Industry
Start with the job your team actually does. Each section below takes the work Alfrada OS is deployed for in that industry — the same jobs listed on strategizelabs.com — and turns it into copy-ready prompts. Replace the bracketed placeholders, attach your files with the + button, and run.
Every prompt works on every plan. Where a job is heavy enough to want parallel workers, the note under the prompt says so — ask for Swarm or let the agent suggest it. Where the job should repeat, the note says daily or weekly — say so and the Task Scheduler takes it from there.
For the fundamentals that make every one of these land better, read Writing Better Prompts first. For prompts organised by tool rather than by job, see the Tool Prompt Library.
Hedge Funds
Coverage that doesn't scale with headcount. Agents read the filings, run the simulations, and draft the memos — your analysts spend the day deciding, not digesting.
Run this in Swarm. Thesis: [one-paragraph thesis on TICKER]. Assign one worker to argue the long case and one to argue the short case, each citing filings, transcripts, and current data. Then have a third worker adjudicate and return: a portfolio recommendation, suggested position size as a % of NAV, the three facts that would change the call, and a one-page memo I can forward.~2 h · Swarm. Pair with Company Financials, Analyst Views, and Stocks Analyser.
Here is my book: [attach holdings CSV]. Build a scenario engine in the sandbox and run: a 20% equity drawdown, a +150bp rate shock, and a stagflation case. For each, show P&L by position, portfolio-level drawdown, and the five assumptions the result is most sensitive to. Return an Excel workbook with one tab per scenario and a summary tab.~1 h · Swarm. Uses Code Executor and Excel Ops.
Read every earnings transcript published this week for [list of tickers]. For each, compare management's commentary to my thesis notes [attach or paste], flag surprises (positive and negative), and rank the names by how much the call moved my conviction. Deliver a table plus a short note per name.~45 min. Set it as a scheduled task during earnings season.
Digest this data room [attach files]. Produce: 1) a comps table against [peer set], 2) a red-flag list with the source document and page for each flag, 3) a first-draft IC memo in our house format [attach template if you have one] covering thesis, valuation, risks, and open diligence questions. Do it in Swarm so the comps, red flags, and memo are worked in parallel.~3 h · Swarm. The team argues with the memo instead of assembling it.
Audit the attached financial model. Walk every formula on the revenue build, flag hard-coded numbers, broken links, and circularities, stress-test the revenue line with a 20% downside, and rank the assumptions by sensitivity to the equity value. Return a findings sheet with cell references, and write the corrections back into a copy of the workbook.~40 min. Excel Ops reads and patches the workbook in place.
Track live job postings for [company] and [peers] and tell me what they say about spending and direction: mix by function, seniority and location, what is new since last run, published pay bands, and the technologies the requirements name. Treat it as a fixed-query sample, not a headcount — and flag reposted evergreen roles rather than counting them as fresh demand. Then schedule it weekly and append to a workbook so I own the trend.~30 min · weekly. The only continuous operating disclosure a private comparable gives you.
Every weekday at 07:00 my time, land a pre-market brief in my inbox on [sectors or tickers]: overnight developments, notable movers, macro prints due today, and what matters for my book. Cite sources. Keep it to one screen.~10 min · daily. Task Scheduler plus Google News and My Gmail.
See also: Finance, Document Analysis, Dashboards.
Strategy & Consulting
The junior half of every engagement, finished by agents. Research, synthesis, and deck production run in parallel while partners edit the argument instead of assembling it.
Run a market entry study for [client] into [market]. In Swarm: size the market top-down and bottom-up, scan the competitor set with positioning and pricing, and synthesise these interview notes [attach]. Then assemble a client-ready deck of 12–15 slides in [our template / a clean consulting style]: executive summary, market sizing, competitive landscape, entry options, recommendation, and next steps.~4 h · Swarm. Presentation Create for the deck, Report Generator for the appendix.
Our client is considering [strategic move]. Map the second-order effects: likely competitor counter-moves, talent and hiring shifts, capital flows, regulatory reactions, and customer behaviour. Score each on likelihood and impact, and return a decision matrix plus the three effects most likely to be underweighted by the board.~1 h. Ask for a Mermaid diagram if you want the causal chain drawn.
Before we recommend entry into [market], tell me who is committing to it. Search live postings for the roles that market requires across [region]: which companies are hiring, in which cities, at what pay, and when each of them made their first hire there. Rising demand with nobody staffing is an opening; rising demand with three rivals hiring is a contested market — tell me which this is.~30 min. Dates a rival's entry months before the launch announcement.
Compare [Target A], [Target B], and [Target C] as acquisition candidates for [acquirer]. Build a weighted scoring model on strategic fit, financial profile, integration risk, valuation, and cultural fit — with the weights I can change in Excel — and plot the three on a quadrant map. Recommend one, with the conditions under which the ranking flips.~2 h · Swarm. One worker per target, then a synthesis pass.
Rebuild the monthly board pack from these exports [attach]. Produce the narrative (what changed and why), waterfall charts for revenue and EBITDA bridges, KPI trend charts, and a deck in the attached template. Flag any number that moved more than 10% month-on-month with a one-line explanation.~1 h. Excel Ops plus Presentation Create or PPTX Patch.
Here is the client's deck [attach PPTX]. Update slides [numbers] with the figures in this sheet [attach], keep every design element, layout, and animation exactly as is, and return the patched file plus a list of every change you made.~20 min. PPTX Patch edits in place; nothing is regenerated.
Every Monday at 08:00, send me a digest of the week's developments across [industries you cover]: deals, regulation, leadership moves, and notable data — with citations. One paragraph per industry, then a short list of items worth raising with clients.~15 min · weekly. Task Scheduler with Google News and Tavily Search.
See also: Case Studies, Presentations, Research And Automation.
Marketing & Creative
One brief in, every asset out — sized per channel, on brand, and checked for consistency before you see it.
Here is the campaign brief [attach or paste]. Produce: three copy variants each for LinkedIn, Instagram, and email; two hero images in our visual style [attach references]; and a 20-second cut for Reels and one for YouTube Shorts using this footage [attach]. Deliver each asset sized for its channel, plus a one-page summary of what you made and why.~1 h. Image Lab, Video Lab, and the CUDA Video Editor in one session.
Attached are [N] product photos and our brand guidelines [attach]. Re-edit every photo to the guidelines — background, colour treatment, crop, and margin rules — keep the product itself unchanged, and return them as a zip with a before/after contact sheet.~30 min. Ask for an Image Consistency check if the set must stay uniform.
Make a 30-second product video for [product]: spoken dialogue from this script [paste], a music bed that fits [mood], and on-screen captions. Use the attached product images as reference and run a character-consistency check across the shots before you deliver. Return the final MP4 plus the consistency verdict.~40 min. Video Lab plus Music Lab; Image Consistency audits the frames.
Here is a 20-minute recording [attach or paste the YouTube link]. Cut it into a 60–90 second highlight reel for social: pick the strongest moments, add burned-in subtitles, and export in 9:16 and 16:9. Give me the timestamps you chose and why.~20 min. CUDA Video Editor; YouTube links are ripped straight into the session.
Connect to My GA4 for [property] and pull the last two quarters. Run channel attribution (last-click and data-driven if available), show cost per acquisition by channel using this spend sheet [attach], and propose next quarter's budget reallocation with the expected impact and the assumptions behind it. Publish it as an interactive dashboard.~1 h. My GA4 plus HTML Dashboard.
Every weekday at 09:00, monitor [brand] and [competitors] across X, Reddit, YouTube, TikTok, and Instagram. Report new mentions, sentiment shifts, viral posts, and anything that needs a response — with links. Flag anything urgent immediately rather than waiting for the digest.~10 min · daily. Task Scheduler with the social search tools.
See also: Creative Studio, Canva, Research And Automation.
Sales & Business Development
Show up briefed, answer faster than the competition, and turn every call into pipeline hygiene.
Draft responses to this RFP [attach] using our past answers [attach] and product docs [attach or connect My GDrive]. Fill the client's own template exactly, mark every answer with a confidence level and its source, and list the questions that need a human. Return the completed document plus the open-questions list.~1 h. Doc Patch keeps the client's formatting intact.
I'm meeting [company] on [date]. Give me a one-page brief: org and who I'm meeting, recent news, hiring signals from their live job postings across every board, stated priorities from their site and filings, and three questions likely to open the conversation. Cite everything.~10 min. Google News, LinkedIn Jobs, LinkedIn Profile Lookup, and Browser.
Our ICP: [describe firm size, industry, geography, and roles]. Find [N] matching companies and the right contact at each, enrich them with My Hunter, and put the results in a Google Sheet with company, contact, role, email, source, and a one-line reason they fit. Flag any you're not confident about.~30 min. My Hunter for discovery and enrichment; results land in My GSheets.
Turn these product notes [attach] into a 10-slide deck for [buyer name and role] at [company]. Structure: their pain in their words, our proof (cases and numbers), how it works for their situation, commercial options, and a clear call to action. Use [our template] and keep every claim traceable to the notes.~30 min. Presentation Create, or PPTX Patch if you have a house deck to adapt.
Here is the call transcript [attach or paste]. Return CRM-ready notes: summary, pain points, objections, next steps with owners and deadlines, and the deal stage you'd recommend. Then draft the follow-up email in my voice.~5 min. Add My Gmail to send the follow-up straight from the session.
Every Friday at 16:00, digest my pipeline sheet [connect My GSheets] and this week's competitor news into one page for the Monday sales call: deals that moved, deals that stalled, competitor launches and pricing changes, and three actions for the team.~15 min · weekly. Task Scheduler with My GSheets and Google News.
See also: Workspace & Email, Presentations, Research And Automation.
Insurance
Submissions, claims, and policy wordings digested into structured positions — with every clause traced back to its source.
Digest this underwriting submission [attach]. Produce a risk summary (exposure, loss history, controls, red flags with document and page), the information still missing, and a draft quote sheet in our template [attach] with the fields you can populate and the ones that need an underwriter.~30 min. Doc Patch fills the quote sheet without touching its formatting.
Review this claims file end to end [attach policy, correspondence, and evidence]. Reconstruct the timeline, map the facts to the policy terms and exclusions, identify coverage issues and evidence gaps, and draft the coverage position letter for review. Cite the document and page for every point.~45 min · Swarm for large files. Medical Vision reads clinical documents on your deployment.
From this policy portfolio [attach], extract every obligation, exclusion, endorsement, and notification requirement into one table: policy, section, clause text, effect, and deadline. Return it as an Excel workbook with a tab per policy and a consolidated view.~40 min. Excel Ops builds the workbook.
Compare these two policy wordings [attach both]. Table every changed clause with the old text, the new text, and its significance for coverage. Rank the changes by materiality and write a one-paragraph summary for the broker.~20 min. Works on PDF and Word alike.
Here is our book [attach exposure data]. Model downside scenarios — [cat event], reserve deterioration of [X]%, and a [Y]% rate change — and rank exposures by sensitivity to each. Return an Excel workbook with the model and a summary I can take to the risk committee.~1 h · Swarm. Code Executor for the simulation, Excel Ops for the deliverable.
Every Monday at 08:00, digest regulatory and supervisory changes across [markets and lines] from the past week into one page: what changed, who it affects, effective dates, and what we should do. Cite the source documents.~15 min · weekly. Task Scheduler with Google News and Tavily Search.
See also: Document Analysis, Data, Dashboards.
Data Science & Analytics
Analysis written up in business English, models prototyped in a disposable sandbox, and findings published as dashboards that keep themselves current.
Here is a dataset I haven't seen before [attach]. Run exploratory analysis: schema and data quality, distributions, outliers, correlations, and anything surprising. Write it up in business English with the charts inline, and end with the five questions this data could answer next.~20 min. Code Executor for the analysis, Report Generator for the write-up.
Using this history [attach], prototype a forecasting model for [target] in the sandbox. Compare at least two approaches against a naive baseline, report accuracy on a holdout, show the forecast with confidence bands, and be explicit about what would break it in production.~30 min. Code Executor; ask for Swarm to try more approaches in parallel.
Here is our model and its assumptions [attach]. Run Monte Carlo simulations on [assumptions], with distributions you justify from the data, and hand back confidence bands rather than point estimates for [outputs]. Show which assumptions drive most of the variance.~30 min. Code Executor; results land in Excel if you ask.
Clean and normalize these exports [attach]. Standardise column names, types, dates, and categories, deduplicate, and reconcile IDs across files. Return analysis-ready tables plus a log of every transformation you applied and every row you dropped, with the reason.~20 min. The transformation log is the point — insist on it.
Launch [long-running job] in the sandbox on this data [attach]. Wake yourself to check on it, report progress to me, and back off between checks until it finishes. When it's done, summarise the result and attach the outputs.~hours, unattended. Wake Me handles the check-ins.
Publish the findings as an interactive dashboard: [metrics and views]. It should re-pull its own data from [source] and update in place on a [daily/weekly] schedule, with a note showing when it last refreshed.~30 min. HTML Dashboard plus Task Scheduler.
See also: Data, Dashboards, Coding.
Supply Chain & Logistics
Simulate the disruption before it happens, score suppliers on evidence, and keep watch on the lanes that matter.
Here is our supply network [attach nodes, lanes, lead times, and volumes]. Simulate [supplier failure / port closure at X] and trace the effect through the network: stock-outs by SKU and week, cost to expedite, and service impact. Rank the mitigations by cost and lead time and recommend one.~1 h · Swarm. Code Executor for the simulation.
Score every supplier in this list [attach] head-to-head on price, lead time, reliability, and quality using the attached performance data. Build a weighted scorecard in Excel where I can change the weights, and recommend the sourcing split with the reasoning.~40 min. Excel Ops.
Review the supplier contracts in this folder [attach or connect My GDrive] for expiry dates, renewal notice periods, penalty clauses, and asymmetric obligations. Return one table with contract, counterparty, clause, effect, and date — sorted by what needs action first.~45 min. Works across PDF and Word.
We're considering [a new DC / reshoring a line / a new 3PL]. Map the second-order effects across cost, service, working capital, carbon, and risk, score likelihood and impact, and return a decision matrix plus what we'd need to believe for it to pay back.~1 h.
Scan this operational data [attach] for anomalies and trend breaks in OTIF, inventory turns, and freight cost per unit — by lane, site, and supplier. Rank the findings by business impact and, for each, say what you'd check first.~20 min. Turn it into a live dashboard once the view is right.
Every weekday at 07:30, monitor freight rate movements, port and weather disruptions, and news on our suppliers [list] across our lanes [list]. Report what changed overnight and anything that needs a decision today.~10 min · daily. Task Scheduler with Google News and Tavily Search.
See also: Data, Dashboards, Document Analysis.
Real Estate & Construction
Underwrite faster, extract every lease term, compare bids on evidence, and keep lenders updated without rebuilding the report each month.
Underwrite this deal [attach pro forma and offering memo]. Audit every assumption in the pro forma, stress-test vacancy, rent growth, exit cap, and rate downside scenarios, and return the sensitivity tables plus a one-page view: is the return there, and what has to be true.~45 min. Excel Ops reads and stresses the workbook.
Digest this acquisition data room [attach]. Produce a red-flag list with source and page for each item, and a first-draft investment committee memo covering the asset, market, business plan, returns, risks, and open diligence questions.~2 h · Swarm.
From this lease portfolio [attach], extract every rent step, break clause, renewal option, service charge term, and tenant obligation into one table with lease, tenant, clause, date, and effect. Return an Excel workbook with a tab per lease and a consolidated schedule.~40 min. Excel Ops.
Compare these contractor bids [attach] with a weighted scoring model on price, programme, exclusions, risk allocation, and track record. Normalise the bids so like is compared with like, flag scope gaps, and recommend one with the conditions.~30 min. Weights adjustable in the Excel output.
Turn this month's draw requests and site reports [attach] into a lender-ready update: narrative on progress and variances, cost-to-complete, schedule status, charts, and photos where provided. Use the attached template and keep the format the lender expects.~30 min. Doc Patch or Presentation Create depending on the template.
Build a pipeline and portfolio dashboard from [connect My GSheets or attach]: deals by stage, occupancy, NOI vs budget, debt maturities, and cost-to-complete by project. It should re-pull the data weekly and update in place.~30 min. HTML Dashboard plus Task Scheduler.
See also: Document Analysis, Finance, Dashboards.
Finance & Accounting
Reconcile, normalise, and narrate the close — with every anomaly traced to its source cell.
Reconcile these spreadsheets for [quarter] [attach]. Match balances and transactions across them, flag every anomaly with the file, sheet, and source cell, and classify each as timing, error, or unexplained. Return the reconciliation workbook and a short summary of what needs a human.~30 min. Excel Ops.
Normalize these operational exports [attach] into analysis-ready tables: consistent columns, types, currencies, and periods, with duplicates removed and IDs reconciled. Include a log of every transformation and every dropped row with the reason.~20 min.
Turn this folder of PDF invoices and statements [attach] into one clean, structured ledger: date, counterparty, description, amount, currency, tax, and reference — with a link back to the source page for each line. Return an Excel workbook and flag any document you couldn't read with confidence.~40 min. Excel Ops builds the ledger.
Audit this financial model [attach]. Check formulas and links, list every hard-coded assumption, stress-test the projections with a 20% revenue downside, and rank the assumptions by their effect on the outcome. Return a findings sheet with cell references and the stressed case as a new tab.~40 min. Excel Ops.
Draft the monthly close narrative from these numbers [attach]: what changed and why, QoQ and YoY variance analysis by line, waterfall charts for the revenue and cost bridges, and the items the CFO should know about. Return a Word document in our template [attach] plus the charts as images.~30 min. Doc Patch keeps the template; Report Generator if you'd rather start fresh.
Build an executive KPI dashboard from [connect My GSheets or attach]: revenue, gross margin, cash, burn or free cash flow, DSO/DPO, and headcount — with trend, budget, and prior year. It should re-pull the data on the [1st of the month / every Monday] and update in place.~30 min. HTML Dashboard plus Task Scheduler.
See also: Finance, Data, Document Analysis.