Resource
Missing Data in AI Extraction — Null, Unknown, or Skip
The most consequential setting in any extraction prompt: what the model does when the field isn't in the text. Four behaviors, and when each is right.
Structured Output
Extraction Prompt Generator
Define what the AI should pull out of a text — invoices, emails, résumés, tickets, contracts — and get an extraction prompt with field definitions, name-aware extraction rules, missing-data behavior, and an ambiguity policy.
What should be pulled out of the text, and for what purpose? E.g. "Extract customer information from support emails."
Fields to Extract *
Each field names a piece of information, not a data type — the description tells the model what to look for. Reorder with ↑ ↓.
Suggested for General Text
Extraction Rules Preview (live — derived from your field names)
⚠ Skip Field + CSV would shift columns — the contract will instruct empty cells instead.
Extraction Prompt
Resource
Extract Contract Information with AI
Parties, effective date, term, payment, termination notice, governing law — key terms into a contract register, with "unknown" marking every gap loudly.
Resource
Extract Data From Text with AI
Free text in, named fields out. The extraction prompt pattern that turns any unstructured text into consistent, parseable records.
Resource
Extract Fields From Emails with AI
Sender, company, request, deadline — out of emails with quoted replies and signature blocks, using guidance that knows how email is actually read.
Resource
Extract Invoice Data with AI
Invoice number, vendor, dates, total, currency — extracted into clean fields with strict no-inference rules, ready for accounts payable.
Resource
Extract Action Items From Meeting Notes
Decisions actually made, commitments actually given — extracted from fragmentary meeting notes that never label their action items.
Resource
Extract Product Review Insights with AI
Pros, cons, feature requests, rating — review text into feedback-board fields, with experienced-vs-wished kept strictly apart.
Resource
Extract Resume Information with AI
Candidate name, current role, years, skills, education — résumés into consistent screening records, with inference kept on a short leash.
Resource
Extract Support Ticket Metadata with AI
Product, issue summary, stated severity, steps already tried — ticket fields extracted from free-text customer messages, without the model's own judgment leaking in.
Resource
Information Extraction Prompt — the Anatomy
The six sections a reliable extraction prompt needs: source guidance, field definitions, extraction rules, missing-data behavior, ambiguity policy, example.
Workflow
AI Data Extraction Workflow
Turn messy text into structured data you can trust enough to feed another system — bound the source, extract the fields, force clean JSON, and validate before it flows downstream.
Workflow
AI Classification Workflow
Build a text classification step you can automate on — pull out the unit to classify, assign a label from a fixed set, and validate the label is one you actually allow.
Workflow
AI Meeting Notes Workflow
Turn a meeting transcript into notes people actually use — a faithful summary, the action items pulled out and assigned, and a clean shareable format.
Workflow
AI Customer Feedback Analysis Workflow
Turn a pile of reviews, surveys, or support comments into themes and priorities — extract the real signal, classify it by theme and sentiment, then summarize what's worth acting on.
Workflow
AI Customer Support Workflow
Run inbound support the same way every time — triage and route the ticket, pull the details that matter, draft a reply in a consistent voice, and log the resolution for the record.
Workflow
AI Hiring Workflow
Run hiring the same way for every role — build a reusable job-description template, lay out a consistent screening sequence, and extract structured data from resumes instead of eyeballing each one.
Project
Build an AI Support Agent with AI
The full path to a support agent you can put in front of customers — write its instructions, ground it in your docs, route and handle tickets, then evaluate and cost-control it before it goes live.
Project
Build an AI Document Processing System with AI
The full path to an AI document processing system — define the use case, design the intake pipeline, extract fields from unstructured documents, classify and route them, pin the output contract, evaluate accuracy, then ship it monitored.
Project
Build an AI Content Moderation System with AI
The full path to an AI content moderation system — define the policy and label taxonomy, extract signals from user content, classify it against policy, emit structured decisions, evaluate false positives and negatives, wire enforcement and review queues, review abuse risks, then ship.
Project
Build an AI Research Assistant with AI
The full path to an AI research assistant — define its scope, organize the source corpus, ground responses in references, extract key facts, synthesize findings, check groundedness, then validate it for use.
Project
Build an AI Meeting Assistant with AI
The full path to an AI meeting assistant — define the use case, turn transcripts into structured notes, extract decisions and action items, classify follow-ups, write a shareable summary, evaluate accuracy, then ready it for the team.
Project
Build a RAG System with AI
The full path to a retrieval system that returns grounded answers — understand the corpus, chunk and ground it, extract and classify the metadata, then evaluate that retrieval actually works.
Project
Build a Programmatic SEO Site with AI
The full path to pages that rank at scale, not penalty bait — map the intents, build the data set, structure it, template the page, then QA before publishing hundreds.
Project
Build a Customer Support System with AI
The full path to a support operation, not just a bot — stand up the knowledge base, route the tickets, add the AI agent, integrate your stack, close the feedback loop, evaluate, and deploy.
Project
Build an Applicant Tracking System with AI
The full path to an applicant tracking system — model jobs, candidates, and hiring stages, generate job descriptions and screening prompts, parse résumés into structured data, design the hiring API, set roles, review security, then ship.
Project
Build a Knowledge Base with AI
The full path to knowledge that's findable by people and AI — plan the taxonomy, structure it for search, write the articles, tag the metadata, make it retrievable, then ship it maintainable.
Project
Build a Data Pipeline with AI
The full path to a pipeline that moves data without corrupting it — design the ingestion and transforms, extract and structure the sources, gate the quality, store it, then deliver and ship it monitored.
How it works
Describe the extraction goal, pick the source type — email, invoice, résumé, support ticket, contract, meeting notes, product review, or general text — and define the fields to extract: a name, a required flag, and a description of what information the field holds. The source type changes the prompt's reading guidance and suggests fields you can add with one click. The engine derives name-aware extraction rules automatically — an email field gets "valid email address only", an amount field gets "numeric value only", a date field gets ISO formatting — and you see them live before generating. Choose what happens to missing data (empty, null, unknown, or skip) and how the model should treat ambiguity (strict, conservative, or best guess). Click Generate Extraction Prompt for the full prompt: source guidance, field definitions, extraction rules, missing-data behavior, ambiguity policy, and an example extraction in JSON, YAML, XML, or CSV. Nothing leaves your browser.
Best for
- Pulling defined fields out of unstructured text — emails, invoices, tickets, résumés
- Setting rules for missing data and ambiguous values up front
- Getting consistent extraction across many similar documents
Not for
- Designing the JSON shape of structured output in general — that is the JSON Output Prompt Builder
- Sorting items into a fixed set of labels — that is the Data Classification Prompt
Use cases
- Pulling invoice numbers, totals, and due dates out of emailed invoices
- Turning résumés into screening records with consistent fields
- Extracting ticket metadata from free-text customer messages
- Mining contracts, meeting notes, and reviews for the fields that matter
Pro tips
- Missing-data behavior is the most consequential setting on the page. Pipelines want Return Null (stable keys); spreadsheets want Leave Empty or Return Unknown (visible gaps); Skip Field only suits consumers that tolerate absent keys.
- Field descriptions do the extraction work: "Decisions actually made — not topics discussed" filters better than any rule. Write descriptions that say what does NOT belong in the field.
- Use Strict ambiguity when wrong data is worse than no data (finance, legal), Best Guess when a blank field is worse than an imperfect one (lead enrichment).
- Name fields the way the engine can help: total_amount gets a numeric-only rule, issue_date gets ISO formatting, skills gets list handling. The rules preview shows what each name earns before you generate.
FAQ
How is this different from the JSON Output Prompt Builder?
The JSON tool defines the output's structure — the contract any task returns its data in, with types and strictness levels. This tool defines what to extract: which information to pull from a text, how to read the source, what to do when a value is missing or ambiguous. The output format here is deliberately light; when you need full contract control, generate the extraction here and tighten the format there.
Why don't extraction fields have a type?
Because an extraction field names a piece of information, not a data shape. "total_amount" means "find the grand total in this text" — whether it serializes as a number is a formatting concern. The engine still infers sensible example values (numbers for amounts, lists for skills, true/false for reply_needed), but the field definition stays about meaning.
What does the missing-data setting actually change?
The MISSING DATA block of the prompt. Leave Empty returns an empty string, Return Null keeps the key with null, Return Unknown writes the literal "unknown", Skip Field omits the key. The engine adapts each to the output format honestly — CSV has no null, so cells stay empty; CSV columns can't be skipped, so the contract instructs empty cells instead.
Is classifying text the same as extracting from it?
No, and the boundary matters: pulling a value out of the text (a name, a total, a date) is extraction — this tool. Choosing a label from a closed set you define (spam/not-spam, positive/negative) is classification — that's the Data Classification Prompt in this category. A severity field that copies the customer's own words is extraction; deciding the severity yourself is classification.
What are the extraction rules the preview shows?
Name-aware rules the engine derives per field: email fields get "valid address only", phone fields get normalization, dates get ISO format, amounts get "numeric value only", identifiers get "exactly as written", list-like fields (skills, action_items, pros) get one-entry-per-item handling. Fields without a matching pattern rely on their descriptions — the preview tells you which.
Why does the source type matter?
Because reading an invoice is not reading meeting notes. The source type adds reading guidance to the prompt — "values follow printed labels", "prefer the newest message over quoted history", "action items may be phrased as commitments" — and suggests the fields that source usually yields. It's the difference between a generic scraper and a prompt that knows what it's looking at.