Sapien Amplified in Action

Six complete workflows, from your first question to the record you keep

The following examples show what you supply, what AI examines, where you respond, and what you keep. Each example addresses a specific goal. Choose the amount of scrutiny appropriate to the money, time, and reputation involved.

You run the supplied prompts in your own AI account. Where a second model is used, you copy the specified material to that model. Where evidence is missing, you obtain the record or answer outside the conversation. You make the real-world decision.

All situations, figures, exchanges, and results below are invented for illustration. A sample finding shows one possible outcome of the described evidence. Your own finding depends on your materials and answers.

Scenario A | Decide whether to acquire a competitor or grow your business

AI Decision Alpha | Strategize with Precision

Your goal: Decide whether to buy a smaller competitor or invest the same available capital in your current business.

Your situation: You run a specialist business-services company. The acquisition could add customers and geographic coverage. Organic growth could be funded in stages. You have already discussed the target with AI for two weeks, and the conversation contains forty messages. The seller's growth story relies heavily on customer retention and cross-selling.

Supply your evidence and decision limits

Bring the seller's proposal, target financial information, proposed terms, your organic growth plan, available capital, and decision deadline. State the operating reserve you must retain and the management capacity you can commit. Supply the complete earlier transcript if you want that conversation examined.

Stage 1 | Check the question before evaluating the acquisition

You do: Add the Reasoning Baseline, the instructions for evidence checks and clear reasoning, using the setup appropriate to your account. Run Foundation Audit, the prompt that examines assumptions and framing in your question.

AI does: Examines whether the original question, “Should I buy the competitor to accelerate growth?”, defines the result you actually need. The AI asks about profit, cash reserves, and management capacity before using growth as the deciding measure.

Sample exchange: AI: “What improvement must the acquisition produce, and how much cash must remain available for your existing business?” You: “The acquisition must improve durable profit while preserving the reserve needed to operate through a weak quarter.”

You confirm: Compare the feasible uses of capital against that objective and your stated reserve.

Stage 2 | Find the unsupported estimate in the earlier conversation

You do: Run Full Conversation Audit, the instrument that examines how reasoning developed across the complete exchange. Provide the transcript and ask the AI to identify the messages available for inspection.

Illustrative audit trail: At message 11, AI estimated how many target customers would buy your additional services. You accepted the estimate without a source. From message 14, projections treated that estimate as established. At message 27, you asked whether the estimate had been verified, and the question remained unanswered.

AI does: Identifies message 11 as the earliest correction point and names the profit projections and acquisition recommendation that depend on the estimate.

You do outside the conversation: Obtain the target's actual customer and service-purchase data. Provide the records to AI and have the affected projections rebuilt. An estimate unsupported by those records remains uncertain.

Stage 3 | Compare materially different uses of your capital

You do: Run Solution Spectrum, the prompt for exploring distinct ways to solve the problem.

AI does: Examines a full acquisition, selected asset purchases, a commercial partnership, staged organic expansion, and waiting for decisive evidence. The AI separates alternatives that change the economics or commitment and merges cosmetic variations.

You respond: Identify which options are actually available. If the seller refuses an asset sale, remove that option and explain the constraint. The remaining alternatives become the comparison set.

Stage 4 | Have two models analyze the open decision separately

You do: Run Cross-Model Decision Triangulation, the protocol for separate initial analyses followed by an exchange of critiques. Give both models the corrected question, evidence, feasible alternatives, and constraints. Keep the earlier recommendation out of their initial analysis contexts.

AI does: Each model produces its own recommendation. You then exchange the analyses under the supplied rules. The models identify the assumptions or evidence differences that could change the decision.

Sample exchange: AI: “The acquisition case depends on customers staying after the owner leaves. Which renewal records or commitments support retention?” You: “I have the seller's estimate but no customer-level evidence for the period after departure.”

Stage 5 | Identify the proof that controls the next commitment

AI does: Shows how different retention assumptions affect the acquisition economics and names the records needed to examine those assumptions. The models either agree on the same analysis or record the material differences still open at the three-round cap.

You decide: Request the evidence, negotiate different terms, pursue organic growth, or accept the remaining uncertainty within your own authority. Another model response cannot supply a private record that neither model has received.

Stage 6 | Record your choice and the result you expect

You do: Complete the Decision Record, the written account of your choice, alternatives, reasons, uncertainty, reversal conditions, and review date. Use Decision Optimizer to record the result you expect and the measure you will check later.

AI does: Helps preserve the expected outcome before you act. When you later supply the original record and actual results, Decision Optimizer compares the evidence with your expectation.

Keep the recommendation and its conditions

Illustrative next action: Continue organic growth planning and request the target's retention records before committing to the acquisition.

Reason: The acquisition economics depend materially on customer continuity after the seller exits. The supplied information does not yet substantiate that continuity.

What could change your choice: Supporting customer records, revised terms that reduce the downside, or evidence that organic growth cannot meet your objective.

What you keep: Corrected projections, feasible alternatives, the models' material differences, evidence requests, your Decision Record, and the expectation recorded for later comparison.

What the example demonstrates: You can correct an earlier AI conversation, compare competing uses of capital, and identify the exact proof needed before a commitment. You select the instruments because each addresses a specific problem in the example.

Scenario B | Decide how to raise prices and what evidence to collect first

AI Decision Alpha | Out-Think Competitors

Your goal: Decide whether to raise prices, which customer groups should receive the increase, and what evidence would justify expanding the change.

Your situation: You sell a recurring B2B service. Costs have increased. You are considering a monthly price change from $200 to $230. Your customers differ in usage, delivery cost, and renewal terms. All figures below are hypothetical.

Supply prices, customer groups, and delivery economics

Bring current prices, renewal dates, discounts, delivery costs, retention history, customer feedback, and comparable offers. State whether your goal is higher revenue, higher contribution profit, or relief from unprofitable work. Identify contractual constraints and customer groups available for a limited test.

Stage 1 | Examine the assumption behind a uniform increase

You do: Use the Reasoning Baseline and Foundation Audit on “How do we raise everyone's price by 15% without losing customers?”

AI does: Identifies the assumed uniform increase and the unsupported expectation of no customer loss. The AI asks which business measure matters and whether different customer groups should receive different treatment.

You respond: Confirm that contribution profit is the objective. Ask AI to examine feasible pricing structures and the evidence required for each.

Stage 2 | Compare different pricing approaches

You do: Run Solution Spectrum before selecting a rollout.

AI does: Develops alternatives such as a new-customer price, increases at renewal, usage tiers, different packages, or removal of uneconomic discounts. After the option search, AI compares the feasible approaches using your costs and customer information.

You respond: Identify billing and agreement constraints. Supply missing cost data before treating increased revenue as increased profit.

Stage 3 | Check what the arithmetic establishes

Illustrative calculation: One hundred customers paying $200 generate $20,000 monthly revenue. At $230, 87 customers generate $20,010. In this simplified example, a loss of 13 customers would leave monthly revenue approximately unchanged.

AI does: States that the calculation compares revenue under fixed assumptions. The calculation does not predict cancellations or establish profit. Usage, customer mix, delivery cost, discounts, and timing require separate examination.

Stage 4 | Challenge the recommended test

You do: Run AI Reasoning Audit, the instrument that reconstructs and challenges an AI analysis using the underlying evidence.

Sample exchange: AI: “A test with new prospects can show conversion at the new price. What evidence do you have about existing customers' renewal decisions?” You: “We can test new-customer conversion now. We have not examined renewal behavior at the proposed price.”

AI does: Limits the initial recommendation to the customer group the test can examine. The AI checks whether competitor offers are comparable before using competitor prices to justify your increase.

Stage 5 | Define the action and expectation before changing prices

You decide: Test the new price with eligible new prospects and make a separate decision about existing customers.

AI does: Helps you complete the Decision Record with the action, alternatives, reasons, uncertainty, reversal conditions, and review date. Decision Optimizer records the expected conversion and contribution-profit result, the measures, and the date for comparison.

You approve: The test period, customer group, and evidence needed before expansion. The example provides no universal threshold for a successful pricing change; those conditions depend on your economics.

Stage 6 | Examine actual results before expanding the change

You do outside the conversation: Run the test through your sales process and retain responses and realized economics. Return with the original record, prediction, and results.

AI does: Decision Optimizer examines what the new-customer data supports and what remains unknown about existing customers. The AI identifies whether the original reasoning requires revision.

Keep your pricing recommendation and review conditions

Illustrative next action: Test the higher price with new prospects before changing renewal prices for current customers.

Reason: The first test can examine new-customer conversion and contribution profit. Existing-customer behavior still requires evidence.

What you keep: Compared pricing options, checked arithmetic, an examined recommendation, your Decision Record, and a prediction with measures and a review date.

What the example demonstrates: You can use selected instruments for a focused pricing decision. Additional model analysis is available when the stakes or remaining differences justify more scrutiny.

Scenario C | Decide whether a private-company investment deserves diligence

AI Deal Interrogator | Invest with Conviction

Your goal: Identify whether the opportunity deserves further diligence, what proof you need, and what would make you walk away.

Your situation: You are evaluating a private B2B software business. The materials describe $2 million in annual recurring revenue, rapid growth, and low churn. Revenue appears concentrated among a few customers. All figures and findings in the example are hypothetical.

Supply the investment materials and your criteria

Bring the pitch, financial model, operating records, proposed valuation and terms, your investment criteria, and walk-away conditions. Select the investor role using the Built-In Configuration Fields. Identify the current documents and any figures supplied only verbally.

Stage 1 | Check the numbers before the first interrogation question

You do: Run the Full Interrogation, the prompt that checks numbers, maps the important points, and questions the deal one point at a time.

AI does: Compares the reported growth rate with the revenue series, runway with cash and monthly burn, and market size with its underlying arithmetic. The AI identifies conflicts and names the document or calculation needed to resolve each conflict.

Illustrative number check: The pitch reports eighteen months of runway. The model shows $550,000 in available cash and a constant net burn of $50,000 per month. Simple division gives eleven months. The AI requests the current cash record and monthly cash forecast to determine whether timing, changing burn, or an error explains the difference.

You receive: A specific discrepancy and evidence request before the AI begins the point-by-point examination. The arithmetic does not establish which source is correct.

Stage 2 | Confirm the points that carry the investment case

AI does: Maps recurring revenue quality, retention, concentration, customer acquisition economics, market demand, financing needs, and valuation. The AI identifies the points most capable of changing your next action and asks you to confirm the focus.

You respond: Correct misunderstandings and add any requirement from your own investment criteria. The confirmed focus determines the detailed questions.

Stage 3 | Examine one important point at a time

Sample exchange: AI: “What contracts or billing records support the $2 million recurring-revenue figure, and does the figure include one-time implementation fees?” You: “The spreadsheet includes implementation fees. I do not have a recurring-revenue reconciliation.”

AI does: Records that the stated recurring amount needs proof. The AI identifies the effect on any valuation calculation that uses the full amount as recurring revenue.

You do outside the conversation: Request the reconciliation and underlying records. The AI identifies the request; you obtain the information from the company.

Stage 4 | Examine retention and the other important risks

AI does: Continues through the confirmed points, asking one question at a time. The AI preserves supported evidence, narrows descriptions where the evidence supports less, and records failed points and outside-proof requirements.

You respond: Supply records or identify what remains unknown. Your walk-away conditions continue to matter when a particular answer breaches your limits.

Stage 5 | Request the Deal Sheet and next action

You do: Run the Deal Sheet prompt, which produces the written record of the examination. If you close early, the preliminary sheet identifies unexamined work.

AI does: Opens with what to do next and why, then records supported, weakened, failed, and unresolved points. The AI includes the strongest case for and against the opportunity and the evidence that could change your next action.

Keep the investment finding and evidence requests

Illustrative next action: Pause advancement until the recurring-revenue and runway discrepancies are resolved.

Why: Both figures affect the valuation and the capital the business may need.

Points that held: The supplied customer history supports an existing commercial business.

Points weakened: The reported recurring amount includes revenue needing reclassification. No replacement recurring figure is established yet.

Points failed: No additional point is conclusively failed in the limited evidence described here.

Outside proof needed: The recurring-revenue reconciliation, supporting records, current cash and forecast, and retention evidence for concentrated customers.

Strongest case for: The commercial history could justify further diligence once the important figures are reconciled.

Strongest case against: Unresolved revenue quality and financing needs prevent a reliable assessment of the proposed economics.

What could change your next action: Records that resolve the discrepancies and support economics within your investment criteria.

What the example demonstrates: You receive precise checks and evidence requests tied to the next diligence action. You still obtain the records and decide whether the opportunity deserves more work.

Scenario D | Prepare your fundraising case before investor meetings

AI Deal Interrogator | Raise Smarter

Your goal: Examine revenue projections, market assumptions, and use of funds before presenting your pitch to investors.

Your situation: You are raising capital for a B2B company. Most current business came through your personal relationships. The growth forecast assumes that a new sales channel will reproduce those results. Your pitch describes the sales process as repeatable, although the proposed channel has not been tested.

Supply the pitch and evidence behind the forecast

Bring the financial model, revenue history, customer list, pipeline, channel history, proposed raise, use of funds, and investor type. Select the founder role through the Built-In Configuration Fields. Explain the milestone the financing is intended to fund.

Stage 1 | Check the figures and confirm the examination

You do: Run the Full Interrogation with the founder role selected.

AI does: Checks the available revenue, growth, cash, runway, and market calculations for consistency. The AI then maps the important fundraising points, including commercial proof, the growth mechanism, economics, capital needs, and milestone.

You respond: Confirm the focus and correct missing context about the investors you intend to approach. The finding will address whether your pitch is ready for serious scrutiny.

Stage 2 | Examine the evidence of repeatable sales

Sample exchange: AI: “Which customers were acquired through a channel another salesperson could use without your personal relationships?” You: “Most came from my network. Two came through referrals outside my network. We have not tested the proposed outbound channel.”

AI does: Separates evidence that customers buy the service from the assumption that the proposed channel can acquire similar customers at the projected cost. The AI records that repeatability is unsupported in its current form.

Stage 3 | Follow the assumption through your forecast

AI does: Asks, one question at a time, what supports the acquisition cost, sales-cycle length, spending plan, and milestone. The examination identifies which parts of the forecast depend on untested channel economics.

You respond: Provide each number's source. Identify estimates as estimates. Explain how the proposed financing would generate the evidence needed to assess the channel.

Stage 4 | Repair the pitch using the evidence you can defend

AI does: Preserves supported customer history and identifies the language that needs to be narrowed or substantiated.

Illustrative repair: “Our current customers demonstrate demand for the service. The proposed financing will test whether we can acquire similar customers through the planned sales channel at acceptable economics.”

You decide: Whether that wording accurately describes your business and financing objective. You revise the actual pitch and model and provide any new evidence for examination.

Stage 5 | Close the examination and reassess completed repairs

You do: Request the Deal Sheet. Complete the named repairs and bring the revised materials back with the prior sheet when further examination is needed.

AI does: Produces the next action, reasons, evidence states, strongest opposing cases, and conditions that could change readiness. The supported pitch remains limited to the evidence examined.

Keep your readiness finding and specific repairs

Illustrative next action: Make the specified repairs before serious investor meetings.

Why: The pitch currently describes an untested acquisition channel as a repeatable sales process.

Points that held: Paying customers and the commercial history supported by your records.

Points weakened: The description of repeatability. Your records support founder-led sales, while the proposed channel remains untested.

Points failed: No additional failed point is assumed in this illustration.

Outside proof needed: Actual channel economics and the evidence required to assess the growth forecast.

Strongest case for: The existing customer history supports a credible explanation of demand and a defined next commercial test.

Strongest case against: The forecast relies on acquisition economics the current evidence does not establish.

What could change readiness: Accurate pitch language, transparent assumptions, a financing plan tied to the stated milestone, and relevant supporting evidence.

What you keep: The Deal Sheet, named repairs, proof requests, and the strongest pitch supported by the examination.

What the example demonstrates: You can find weaknesses while you still have time to correct your pitch and prepare for the questions investors may ask.

Scenario E | Develop a digital product from your consulting expertise

AI Monetization Engine | Monetize Decisively

Your goal: Develop a product from expertise you currently sell through consulting and test paid demand before a full build.

Your situation: You advise small B2B service businesses on pricing and profitability. You have original worksheets, permitted examples, repeated client questions, and direct relationships with possible customers. You are considering a course, workbook, template collection, or another product.

Supply your assets, customer evidence, and capacity

Bring the material you have the right to use, evidence of problems clients paid to solve, access to prospective customers, and time available. Identify confidential or client-owned material that must be excluded. An asset's origin alone does not establish your right to sell the asset.

Stage 1 | Use Conversation 1 to select a product

You do: Run Find What to Sell, the prompt that develops and compares opportunities from your assets or a market you name.

AI does: Examines different products against reach, ownership, delivery, economics, and seven commercial criteria. In this illustration, the selected product is a pricing decision workbook with worked examples. The AI explains why the workbook fits the repeated problem and your delivery capacity better than the other candidates.

You receive: The Product Sheet, the written record of Conversation 1. The sheet contains the selected product, intended customer, problem, contents, evidence, assumptions, price range, reach, delivery requirements, and the alternative in full.

Stage 2 | Have a second AI examine the selection

You do: Open Conversation 2, preferably with a second AI model. Paste the Product Sheet and run Stress-Test the Selected Product. The second model checks the selection and questions the commercial points one at a time.

Sample exchange: AI: “Your clients paid for advice. What shows they can complete the workbook without your interpretation?” You: “The worksheets need explanation. I can add worked examples, but independent use has not been tested.”

AI does: Identifies the delivery assumption that needs repair. You narrow the promise and add the proposed example before requesting the finding, the written conclusion on whether the product is ready for a buyer test. Advance when the finding is READY FOR BUYER TEST.

Stage 3 | Review the offer before AI writes the marketing material

AI does: Defines the customer, paid problem, contents, exclusions, delivery, objections, and price to test. AI stops for your corrections.

Illustrative offer: “A pricing decision workbook for owners of small B2B service businesses, with guided exercises and worked examples for comparing current and proposed pricing.”

You review: Confirm the promise and delivery date you can honor. In this hypothetical example, you select a $149 test price. The price is an invented test parameter.

Stage 4 | Choose a reachable channel and approve the message

AI does: Selects direct email to relevant existing contacts based on the reach you supplied. The AI writes the email and follow-up and stops for your corrections.

Sample message: “I'm developing a workbook for service-business owners who need to compare a proposed price change with current economics. The first version includes the workbook and worked examples for $149.”

You review: Add the exact contents and delivery date. Correct any promise or example you cannot substantiate. You send the approved material yourself.

Stage 5 | Set the test rules before asking people to buy

AI does: Produces the Offer and Real Buyer Test Plan, the document containing your offer, marketing copy, intended buyers, timing, price, response measures, and next-action rules.

Illustrative test: Present the approved offer to 20 relevant contacts over fourteen days. Three or more net paid orders justify the first defined delivery. One or two justify one specified clarification and a retest. Zero leads to reconsidering the alternative after checking test validity. These are invented conditions for this example.

You do outside the conversation: Send the material, answer questions, handle purchases, and record payments, refunds, and cancellations. Record compliments and requests separately as market signals.

Stage 6 | Read the results and build the digital product

Illustrative result: Four paid orders and one refund leave three net paid orders. The supplied response record supports that the intended contacts received and understood the offer.

You do: Return with the original plan and results and run Read the Results.

AI does: Checks validity, compares three net orders with the prewritten conditions, and gives a LAUNCH finding for the defined first delivery. The AI supplies a 30-day selling plan. The finding supports that next commitment; repeatable demand and larger-scale economics remain to be examined.

You build: Use the AI Digital Product Build Guide, the included instructions for developing and checking suitable digital product files. Supply the exact purchased offer, contents, source material, and requirements. Inspect the files and complete the necessary acceptance checks before delivery.

Keep the commercial plan and the product files

You keep the Product Sheet, second-model finding, offer, email and follow-up, test plan, results, next action, and 30-day plan. You also keep the product files developed through the Build Guide and your checks of those files.

What the example demonstrates: You can move from existing expertise to a specific offer and paid test, then use the included construction guidance to develop the digital product you promised.

Scenario F | Find a paid offer in a restaurant market you understand

AI Monetization Engine | Launch Faster

Your goal: Find a product opportunity in a familiar market, compare possible offers, and test a selected offer with people you can reach.

Your situation: You ran purchasing for twelve restaurants for nine years and now work independently. You understand food-cost problems but have no product. Your possible reach includes 600 subscribers to a newsletter last sent two years ago, 900 LinkedIn connections, and a food distributor's sales representative you know. All quantities and results below are hypothetical.

Supply your market knowledge and actual routes to customers

Name the restaurant types you understand, relevant experience, capabilities, time, and delivery limits. Describe the age and relevance of your contacts. Ask the representative whether introductions would be possible before treating those relationships as available reach. Provide research sources or use an AI account capable of searching.

Stage 1 | Research existing offers in Conversation 1

You do: Run Find What to Sell with the restaurant market and your capabilities.

AI does: Examines available food-cost offers, prices, packaging, customer descriptions, and possible gaps. The AI labels competitor prices and customer counts as market signals. Those sources can indicate a problem to investigate, while demand for your offer remains untested.

Stage 2 | Compare offers you can actually deliver

AI does: Compares a fixed-price food-cost audit, supplier renegotiation service, spreadsheet toolkit, and paid workshop. In the illustration, renegotiation exceeds your delivery capacity at the proposed price and is removed. The AI selects the audit and records the spreadsheet toolkit as the alternative.

You receive: The Product Sheet before any second-model examination. The sheet states what the audit delivers, who could buy, the supporting market signals, delivery economics, proposed reach, and the toolkit alternative in full.

Stage 3 | Use Conversation 2 to examine reach and the promise

You do: Paste the Product Sheet into a fresh conversation, preferably with a second AI model, and run the selected-product examination.

Sample exchange: AI: “When did you last contact the 600 subscribers, and how many are current restaurant owners?” You: “The last newsletter was two years ago. The distributor representative knows active owners and has agreed to make relevant introductions.”

AI does: Questions the usefulness of the older list. The representative's confirmed introductions offer a more direct route to the intended customers. The decision depends on relevant access, and the large subscriber count alone does not establish that access.

AI examines the promise: The audit can identify purchasing and food-cost issues. No evidence supports a promised savings amount. You narrow the offer accordingly and request the finding. Proceed after a READY FOR BUYER TEST finding on the described offer.

Stage 4 | Approve the offer and partner messages

AI does: Writes the audit offer and stops for corrections. The AI then selects the agreed introduction route and writes the partner message and owner email. You review those messages before the test is designed.

Illustrative offer: “A food-cost audit for an independent restaurant, examining purchasing records and food-cost information to identify issues worth investigating and actions to consider.”

You confirm: The required customer records, deliverable, exclusions, delivery date, and representative's participation. You perform the assessment work promised in the service offer.

Stage 5 | Write the conditions and run the first test

Illustrative test: The representative introduces you to 30 relevant owners. You offer a $900 audit over thirty days. After a valid test, three or more net paid bookings mean LAUNCH for the defined delivery. One or two mean change one element and test again. Zero means assess the recorded alternative. All figures are invented test parameters.

You do outside the conversation: Arrange the introductions, send the approved offer, answer questions, and record commercial responses. Keep evidence of whether the intended owners received an understandable offer and could purchase.

Stage 6 | Choose the next test from the actual evidence

Illustrative result: One owner books and pays. The response record supports that the thirty introductions occurred, the intended owners received the clear offer, and the buying process and deadline were usable. The validity check passes in this hypothetical case.

AI does: Compares one net booking with the original rules and recommends changing one element. The AI states that one purchase establishes limited paid interest. The result does not establish whether price, channel, or wording explains the other decisions.

Concrete next test: Ask a small group of non-purchasing owners to explain what they expected to receive from the audit. Use those responses to choose the next change. If the responses reveal confusion about the deliverable, test a clearer description while retaining the price and channel. If another obstacle is supported, the AI should explain why a different change is more informative.

You approve: The selected variable, reason for changing that variable, revised offer, and new result conditions before another paid test. No automatic price reduction is assumed.

Keep the finding and the alternative ready for examination

You keep the Product Sheet, finding, offer, partner message, owner email, original test plan, response record, and the reason for the next test. If the second valid test meets its prewritten USE THE RUNNER-UP condition, retrieve the spreadsheet toolkit from the Product Sheet and have a fresh AI conversation examine that product before building.

What the example demonstrates: You can begin with industry knowledge, compare services and products, distinguish a large old contact list from confirmed customer access, and decide the next commitment using actual responses. A service still requires you to deliver the promised service.