Explore the massive dataset everyone has ignored.
Your company may already have the data for a useful discovery. It could be buried in years of service records, a rarely opened operational database or a spreadsheet that somebody dutifully updates and nobody discusses. Let AI find out what's available, look for the neglected corners and investigate what they can tell you.
Try it now
Copy the prompt into your AI tool of choice, then answer its questions as they come.
Paste it into any AI chat to start. For most people, this is all you need.
Everything on this page in one paste: the prompt plus the full plan, every step and what to check. Use it for longer moves, to plan ahead, to share with a colleague, or if your AI can't open links.
How it goes
What the AI will do
- Discover which datasets and records you can actually access, then compare them with what people already discuss and report.
- Explore promising overlooked data, investigate unexpected patterns and verify the strongest findings against the underlying records.
What to have ready
- Your role, team or business area; a rough starting point is enough.
- Access to available data and the reports, correspondence and meeting records that show how it is currently used.
Where you come in
- Clarify any important gaps in what the AI can access or understand.
- Choose the promising area to investigate and help interpret discoveries that need business knowledge.
What you'll get
A sourced investigation of a neglected dataset: useful findings, their limits, reproducible analysis and concrete questions or decisions they can inform.
You're done when
You have investigated a promising overlooked dataset and received traceable findings, reproducible analysis and a clear account of what was and wasn't examined. Findings that did not survive checking are identified as such.
These are general instructions, not a description of this user's particular AI system, tools, access, or circumstances. Adapt the method to the user's goal and the capabilities you can actually use, while preserving the activity's essential sequence, review points, and intended result. ### Start with the activity and the user's context Read the complete activity and guide the user through it. Use the activity prompt yourself; the user does not need to paste it again. Keep the responsibilities of the user, the AI, and any colleague clear. Begin with what the user has already supplied. Establish the intended result and resolve missing context through focused questions when needed. Work within the activity's stated scope, permissions, required confirmations, and stopping points, including any deliberate order for introducing information. Make progress with the available information and reasonable, clearly identified assumptions where the activity allows them. ### Check the capabilities needed for this work Inspect relevant available tools, connectors, and data. Use small, relevant, non-destructive checks to establish whether necessary access or a capability actually works. Naming a product does not establish access to it. Let this activity determine what needs checking; do not impose a universal preliminary questionnaire or a full system audit. When useful access is missing: - Explain what you cannot reach and how that material would help. - Investigate practical setup, connection, export, or other access alternatives before asking the user to gather material or troubleshoot. - If progress depends on access, request only the specific user action or permission needed to enable it or provide the material another way. Continue useful independent work while awaiting the response. - Offer to walk the user through an export or upload. Welcome files, links, copied text, rough notes, or an interview as supported by your system and the activity, and use the best available route. If the missing access is optional, continue with what is available. ### Ask focused questions and respect the answers Ask questions only when necessary to complete the work well or satisfy the activity's required review points. Build on answers already supplied. Ask one question at a time, or a small group of closely related questions. Wait on work that depends on the answer, and continue useful work that does not. Explain unfamiliar language with examples where helpful. Confirm answers or decisions you supply on the user's behalf before treating them as the user's final answers or decisions. This does not require separate approval of every sourced finding; follow the activity's specific review requirements. ### Keep evidence and independent contributions distinct Distinguish evidence, inference, and missing information. Check your work against the source material. Use independent agents when your system actually supports delegation and independent investigation or review would improve the work: - Give each agent a bounded brief and the context appropriate to its role. - Keep discovery independent until synthesis where the activity requires it. Preserve any requirement for the user or colleagues to contribute before seeing other explanations. - Follow the activity's stated fallback and any required confirmation when independent agents or fresh contexts are unavailable. Preserve reviewer isolation where required; describe shared-context or sequential passes accurately. ### Produce, check, and revise the result Do the substantive investigation and production. Carry the activity through to the most complete, usable result within its scope, rather than ending with a plan or handing the main work back to the user. When missing information does not prevent useful work, proceed with clearly identified assumptions where the activity permits them. Check the result against the activity's completion criteria. At each substantive delivery, state what is complete and any material unfinished work, assumptions, helpful missing information, and remaining uncertainty. Explain how these affect what the user can use, and help the user revise the result. Preserve the activity's stopping point. Do not turn a rehearsal, investigation, or test-design activity into execution that the user has not authorized. Help me discover useful information hiding in data we have overlooked. Begin by finding out what data is available to me and what you can actually examine. Carry this through to a completed investigation of an agreed, promising area. ### Discover the available data and its current use Use what I have supplied to identify my role, business area, projects and useful search terms. Inspect the available connections, files and company systems. Try relevant access and enumerate accessible data catalogs, databases, spreadsheets, operational records, product or service logs, customer records and archived datasets. Start with metadata and manageable samples; record what exists, what I appear able to access, what you successfully read, and what remains unverified. Don't mistake a connector's name or a catalog listing for access to its contents. Search relevant email, human chat including WhatsApp where accessible, meeting transcripts, reports, dashboards, project documents and calendar context. Look for recurring questions, decisions, named measures, dataset references, prior analyses and known data-quality problems. Use this evidence to understand what the business cares about and which available data receives little attention. Search the data's names, owners, subject terms and aliases; a dataset might be discussed under another name. Show the limits of this comparison: absence from the records searched is evidence of low visibility there, not proof that nobody has used it. Keep a compact inventory of sources, owners where known, time coverage, volume, units of observation, apparent business purpose, access and observed usage. Group large inventories into manageable areas. If access fails, investigate available connection or export routes before asking me for the specific action only I can take. Accept a permitted sample or existing export. If the necessary data is unavailable, explain the narrower investigation possible and get my agreement; an inventory or access request alone is not a completed discovery. ### Choose the neglected area worth exploring Compare promising datasets on apparent neglect, business relevance, information quality and tractability. Explain a few concrete opportunities: for example, recurring service visits might reveal an avoidable second visit, or small order changes might expose a demand signal missing from the standard report. These are hypotheses until tested. Recommend a bounded first investigation, with the data, date range, useful business question and reason to expect something new. Ask me one focused question where my judgment changes the choice and wait for agreement on the investigation. Don't require me to invent the question before you have inspected the landscape. ### Understand and explore the records Inspect definitions, collection history, schema, missingness, duplicates, units, timestamps, changes in systems and how records join. Establish what each row represents. Reconcile representative records and aggregates with their source. Preserve the original data and document cleaning or exclusions. Distinguish a true zero from a missing value and an event time from an entry or update time. Explore distributions, trends, cohorts, unusual combinations, exceptions, sequences and plausible joins that could expose useful information. Check whether patterns change across periods, locations, products or customer groups where relevant. Keep an investigation log of questions, analyses, results and rejected explanations. Use code or appropriate data tools for calculations; don't infer numerical results from a partial text preview. For a large corpus, divide inspection into bounded assignments by dataset, period or analytical question. Where useful and available, give separate agents different exploratory questions with source references, definitions, deliverables and coverage limits. Have them return their findings before reconciling them. Use sequential passes or explicit fresh-session handoffs when agents or capacity are limited. Track exactly what was examined. ### Challenge and verify the strongest discoveries Select the few findings with the strongest combination of evidence and practical relevance. Reproduce their calculations, trace examples to raw records and test alternatives such as seasonality, inconsistent logging, changes in coverage, sample size, outliers or confounding business differences. Where possible, check promising patterns in a period or subset that was not used to discover them. Clearly separate exploratory patterns, replicated findings and causal explanations that the data cannot establish. Use an independent reviewer where available, supplying the analysis, source records and claimed findings without your defense of them. Ask it to reproduce key results and seek explanations that would change the conclusion. Otherwise perform a distinct verification pass and preserve its limits. Resolve material discrepancies before presenting a discovery as established. Ask me about business meaning only after doing the investigation the records permit, then test any new explanation against available evidence. ### Deliver what was hiding and why it matters Provide a concise account of the most useful discoveries with concrete examples, relevant quantities, source links and the business questions or decisions each informs. For each, explain what is established, what remains uncertain and what additional observation would discriminate between competing explanations. Don't manufacture a surprising result: a well-supported finding that the promising pattern disappears is a valid outcome. Deliver the analysis files or reproducible queries, data definitions, transformation notes and an accessible summary. Include the available-data inventory, the evidence for choosing this neglected area and actual investigation coverage. Preserve unresolved promising questions separately from findings. Finish with a completed investigation that another person can inspect and repeat. These instructions were written in September 2026 using the frontier models available at the time. You may now have better tools or models, or restrictions that make some tasks harder. Your AI should adapt to the tools and access you actually have and find the best available way to help. Everything here is a starting suggestion for you and your AI, not a proven best practice for every setup. Feel free to rewrite any of it to better meet your needs.
View full AI instructions
Explore the massive dataset everyone has ignored.
What the AI will do
- Discover which datasets and records you can actually access, then compare them with what people already discuss and report.
- Explore promising overlooked data, investigate unexpected patterns and verify the strongest findings against the underlying records.
What you can provide
- Your role, team or business area; a rough starting point is enough.
- Access to available data and the reports, correspondence and meeting records that show how it is currently used.
Where you'll take part
- Clarify any important gaps in what the AI can access or understand.
- Choose the promising area to investigate and help interpret discoveries that need business knowledge.
What you'll get
A sourced investigation of a neglected dataset: useful findings, their limits, reproducible analysis and concrete questions or decisions they can inform.
Activity context
Your company may already have the data for a useful discovery. It could be buried in years of service records, a rarely opened operational database or a spreadsheet that somebody dutifully updates and nobody discusses. Let AI find out what's available, look for the neglected corners and investigate what they can tell you.
For the user
Start with your business area or an example of data you suspect is underused. You don't need to locate the perfect dataset first. The AI will inspect what is available and recommend where to look. You'll help choose the investigation and interpret findings; it will do the data work and verification.
Guidance for the AI
These are general instructions, not a description of this user's particular AI system, tools, access, or circumstances. Adapt the method to the user's goal and the capabilities you can actually use, while preserving the activity's essential sequence, review points, and intended result.
Start with the activity and the user's context
Read the complete activity and guide the user through it. Use the activity prompt yourself; the user does not need to paste it again. Keep the responsibilities of the user, the AI, and any colleague clear.
Begin with what the user has already supplied. Establish the intended result and resolve missing context through focused questions when needed. Work within the activity's stated scope, permissions, required confirmations, and stopping points, including any deliberate order for introducing information. Make progress with the available information and reasonable, clearly identified assumptions where the activity allows them.
Check the capabilities needed for this work
Inspect relevant available tools, connectors, and data. Use small, relevant, non-destructive checks to establish whether necessary access or a capability actually works. Naming a product does not establish access to it. Let this activity determine what needs checking; do not impose a universal preliminary questionnaire or a full system audit.
When useful access is missing:
- Explain what you cannot reach and how that material would help.
- Investigate practical setup, connection, export, or other access alternatives before asking the user to gather material or troubleshoot.
- If progress depends on access, request only the specific user action or permission needed to enable it or provide the material another way. Continue useful independent work while awaiting the response.
- Offer to walk the user through an export or upload. Welcome files, links, copied text, rough notes, or an interview as supported by your system and the activity, and use the best available route. If the missing access is optional, continue with what is available.
Ask focused questions and respect the answers
Ask questions only when necessary to complete the work well or satisfy the activity's required review points. Build on answers already supplied. Ask one question at a time, or a small group of closely related questions. Wait on work that depends on the answer, and continue useful work that does not. Explain unfamiliar language with examples where helpful.
Confirm answers or decisions you supply on the user's behalf before treating them as the user's final answers or decisions. This does not require separate approval of every sourced finding; follow the activity's specific review requirements.
Keep evidence and independent contributions distinct
Distinguish evidence, inference, and missing information. Check your work against the source material.
Use independent agents when your system actually supports delegation and independent investigation or review would improve the work:
- Give each agent a bounded brief and the context appropriate to its role.
- Keep discovery independent until synthesis where the activity requires it. Preserve any requirement for the user or colleagues to contribute before seeing other explanations.
- Follow the activity's stated fallback and any required confirmation when independent agents or fresh contexts are unavailable. Preserve reviewer isolation where required; describe shared-context or sequential passes accurately.
Produce, check, and revise the result
Do the substantive investigation and production. Carry the activity through to the most complete, usable result within its scope, rather than ending with a plan or handing the main work back to the user. When missing information does not prevent useful work, proceed with clearly identified assumptions where the activity permits them.
Check the result against the activity's completion criteria. At each substantive delivery, state what is complete and any material unfinished work, assumptions, helpful missing information, and remaining uncertainty. Explain how these affect what the user can use, and help the user revise the result.
Preserve the activity's stopping point. Do not turn a rehearsal, investigation, or test-design activity into execution that the user has not authorized.
Your AI prompt
Help me discover useful information hiding in data we have overlooked. Begin by finding out what data is available to me and what you can actually examine. Carry this through to a completed investigation of an agreed, promising area.
Discover the available data and its current use
Use what I have supplied to identify my role, business area, projects and useful search terms. Inspect the available connections, files and company systems. Try relevant access and enumerate accessible data catalogs, databases, spreadsheets, operational records, product or service logs, customer records and archived datasets. Start with metadata and manageable samples; record what exists, what I appear able to access, what you successfully read, and what remains unverified. Don't mistake a connector's name or a catalog listing for access to its contents.
Search relevant email, human chat including WhatsApp where accessible, meeting transcripts, reports, dashboards, project documents and calendar context. Look for recurring questions, decisions, named measures, dataset references, prior analyses and known data-quality problems. Use this evidence to understand what the business cares about and which available data receives little attention. Search the data's names, owners, subject terms and aliases; a dataset might be discussed under another name. Show the limits of this comparison: absence from the records searched is evidence of low visibility there, not proof that nobody has used it.
Keep a compact inventory of sources, owners where known, time coverage, volume, units of observation, apparent business purpose, access and observed usage. Group large inventories into manageable areas. If access fails, investigate available connection or export routes before asking me for the specific action only I can take. Accept a permitted sample or existing export. If the necessary data is unavailable, explain the narrower investigation possible and get my agreement; an inventory or access request alone is not a completed discovery.
Choose the neglected area worth exploring
Compare promising datasets on apparent neglect, business relevance, information quality and tractability. Explain a few concrete opportunities: for example, recurring service visits might reveal an avoidable second visit, or small order changes might expose a demand signal missing from the standard report. These are hypotheses until tested.
Recommend a bounded first investigation, with the data, date range, useful business question and reason to expect something new. Ask me one focused question where my judgment changes the choice and wait for agreement on the investigation. Don't require me to invent the question before you have inspected the landscape.
Understand and explore the records
Inspect definitions, collection history, schema, missingness, duplicates, units, timestamps, changes in systems and how records join. Establish what each row represents. Reconcile representative records and aggregates with their source. Preserve the original data and document cleaning or exclusions. Distinguish a true zero from a missing value and an event time from an entry or update time.
Explore distributions, trends, cohorts, unusual combinations, exceptions, sequences and plausible joins that could expose useful information. Check whether patterns change across periods, locations, products or customer groups where relevant. Keep an investigation log of questions, analyses, results and rejected explanations. Use code or appropriate data tools for calculations; don't infer numerical results from a partial text preview.
For a large corpus, divide inspection into bounded assignments by dataset, period or analytical question. Where useful and available, give separate agents different exploratory questions with source references, definitions, deliverables and coverage limits. Have them return their findings before reconciling them. Use sequential passes or explicit fresh-session handoffs when agents or capacity are limited. Track exactly what was examined.
Challenge and verify the strongest discoveries
Select the few findings with the strongest combination of evidence and practical relevance. Reproduce their calculations, trace examples to raw records and test alternatives such as seasonality, inconsistent logging, changes in coverage, sample size, outliers or confounding business differences. Where possible, check promising patterns in a period or subset that was not used to discover them. Clearly separate exploratory patterns, replicated findings and causal explanations that the data cannot establish.
Use an independent reviewer where available, supplying the analysis, source records and claimed findings without your defense of them. Ask it to reproduce key results and seek explanations that would change the conclusion. Otherwise perform a distinct verification pass and preserve its limits. Resolve material discrepancies before presenting a discovery as established. Ask me about business meaning only after doing the investigation the records permit, then test any new explanation against available evidence.
Deliver what was hiding and why it matters
Provide a concise account of the most useful discoveries with concrete examples, relevant quantities, source links and the business questions or decisions each informs. For each, explain what is established, what remains uncertain and what additional observation would discriminate between competing explanations. Don't manufacture a surprising result: a well-supported finding that the promising pattern disappears is a valid outcome.
Deliver the analysis files or reproducible queries, data definitions, transformation notes and an accessible summary. Include the available-data inventory, the evidence for choosing this neglected area and actual investigation coverage. Preserve unresolved promising questions separately from findings. Finish with a completed investigation that another person can inspect and repeat.
Done when
You have investigated a promising overlooked dataset and received traceable findings, reproducible analysis and a clear account of what was and wasn't examined. Findings that did not survive checking are identified as such.
These instructions were written in September 2026 using the frontier models available at the time. You may now have better tools or models, or restrictions that make some tasks harder. Your AI should adapt to the tools and access you actually have and find the best available way to help. Everything here is a starting suggestion for you and your AI, not a proven best practice for every setup. Feel free to rewrite any of it to better meet your needs.
These instructions were written in September 2026 using the frontier models available at the time. You may now have better tools or models, or restrictions that make some tasks harder. Your AI should adapt to the tools and access you actually have and find the best available way to help. Everything here is a starting suggestion for you and your AI, not a proven best practice for every setup. Feel free to rewrite any of it to better meet your needs.